Characterization of the New Pentafluorosulfanyl-Substituted Chalcone 246TMP-3SF5 as Potential New Treatment Option Against Hepatocellular Carcinoma

1. Introduction

Hepatocellular carcinoma (HCC) is a highly relevant cancer entity as it represents the most common primary liver cancer and accounts for the third highest cancer death rate worldwide (Vogel u. a. 2022; Bray u. a. 2024). Major risk factors for developing HCC include chronic hepatitis infection, excessive alcohol consumption and metabolic disease leading to non-alcoholic-fatty-liver disease (Vogel u. a. 2022; Singal u. a. 2023; Moris u. a. 2025). Often HCC remains asymptomatic for a long period of time with symptoms only appearing later on in progressed illness, therefore often leading to a late diagnosis in advanced stages of HCC (Lim u. a. 2023; Benson u. a. 2021). International therapeutic approaches are based on HCC tumor stages classified by the Barcelona Clinic Liver Cancer (BCLC) staging system depending on tumor size and spread, liver function and physical status of the patient. While early and intermediate stage HCC can be treated by a variety of therapeutic options, such as surgical resection, liver transplantation, local ablation or arterial chemoembolization, the most common diagnosed advanced-stage HCC is commonly treated with a palliative approach using systemic therapies (Dimitroulis u. a. 2017; Reig u. a. 2025; Moris u. a. 2025). First-line treatment consists of immune checkpoint inhibitors and the antiangiogenic antibody bevacizumab in varying combinations, while multikinase inhibitors, such as sorafenib, are also widely used (Vogel u. a. 2022; Singal u. a. 2023; Reig u. a. 2025; Moris u. a. 2025). However, late diagnosis and underlying chronic liver dysfunction restrict therapy outcomes and resistance of tumors to systemic therapies raises many challenges in the treatment of advanced-stage HCC (Chen u. a. 2019). Therefore, advanced-stage HCC is still characterized by high lethality and recurrence rate and consequently poor overall prognosis for patients, with median overall survival of only 12–19 months (Vogel u. a. 2022; Singal u. a. 2023; Reig u. a. 2025; Ivanics u. a. 2022; Tampaki u. a. 2015; Nakamura u. a. 2023). Despite the emergence of targeted therapy agents treatment of advanced-stage HCC remains insufficient challenged by low response rate and resistance to therapy resulting in far from satisfactory therapy results (Chen u. a. 2019). Hence, the development of highly effective and survival prolonging medication against HCC in advanced stages is highly needed.

Heat shock proteins (HSPs), a family of highly conserved proteins also known as molecular chaperons, are emerging targets for new therapeutic agents. HSPs are ubiquitously expressed and are classified and divided into six major families according to their molecular weight (HSP110, HSP90, HSP70, HSP60, HSP40 and small HSPs). They can be upregulated under conditions of cellular stress (e.g., hypoxia, heat shock, genotoxic agents, nutrition starvation), performing a multitude of functions, such as stabilizing proteins, inhibiting stress-induced protein aggregation and regulating cell signaling and transcriptions networks (C. Wang u. a. 2016; Somu u. a. 2024). Malignant cells seem to be more dependent on the chaperone activity of the heat shock protein family because of their high burden of misfolded and mutated proteins, as well as higher metabolic needs. To facilitate rapid cancer cell proliferation, they exploit these cellular pathways, ensuring cancer cell survival and contributing to tumor progression. Correspondingly, overexpression of HSPs has been observed across multiple cancer entities and linked to therapy resistance, poor prognosis and reduced survival, making them a promising target for cancer therapy (Somu u. a. 2024; „Heat Shock Proteins (HSPs) as Chaperones for Oncogenesis” 2025; Chatterjee und Burns 2017; Kunachowicz u. a. 2024).

Particularly, heat shock protein 90 (HSP90), with a molecular mass of 90 kDa, is of great interest because of its central role regulating over 200 client proteins involved in all stages of carcinogenesis. Its client proteins include a wide range of oncoproteins such as kinases (e.g., AKT, RAF), transcriptions factors (e.g., HIF-1α) as well as hormone receptors, all critical for the growth and survival of cancers cells. Moreover, HSP90 plays a critical role in promoting tumor angiogenesis (Meng u. a. 2017). HSP90 has been shown to be upregulated in HCC and correlates with a poor prognosis, making it a promising target for new therapeutic approaches, especially in advanced-stage HCC (C. Wang u. a. 2016; Zuo u. a. 2024; Paul u. a. 2024; Sun u. a. 2024; Nouri-Vaskeh u. a. 2020). Several HSP90 inhibitors have been developed and tested in various cancer entities, the geldanamycin derivative 17-AAG being one of its prominent examples, inhibiting the function of HSP90 by binding to its ATP-binding pocket leading to the degradation of various client proteins. Although preclinical studies showed good effectiveness of 17-AAG, its clinical translation was highly challenged by poor solubility and oral bioavailability as well as hepatotoxic side effects (Chatterjee und Burns 2017; Talaei u. a. 2019). Hence, HSP90 inhibitors display a promising target in targeted therapy against HCC but novel inhibitors need to be developed and evaluated.

Previously, the antiproliferative efficacy of the novel compound 246TMP-3SF5 against HCC cell lines was reported (Viperino u. a. 2025). The drug is derived from the small molecule SU086, an HSP90 inhibitor shown to be highly effective in prostate cancer (Rice u. a. 2022). 246TMP-3SF5 contains a chalcone backbone (2,4,6-trimethoxyphenyl moiety) with a 3-SF₅-substituted phenyl group. Chalcones represent a privileged structure in medicinal chemistry, widely used as a promising chemical scaffold for drugs against various disorders, especially against various cancers (Rajendran u. a. 2022). Both natural and synthetic chalcones have been extensively investigated for anticancer effectiveness and demonstrated promising results against various cancer types including colon, breast, lung and prostate cancer (Shukla u. a. 2021; Ouyang u. a. 2021; Constantinescu und Lungu 2021). Mechanisms of anticancer activity of chalcones include apoptosis induction, inhibition of tubulin polymerization, as well as anti-angiogenic and anti-inflammatory effects by targeting multiple cellular molecules, such as MDM2/p53, tubulin, NF-κB, VEGF, EGFR, HIF-1 and many more (Ouyang u. a. 2021). Modulating the basic structure of chalcones allows for a synthesis of new compounds with improved anticancer activity (Ouyang u. a. 2021; Constantinescu und Lungu 2021). The combination of a 2,4,6-trimethoxyphenyl moiety with a 3-SF₅-substituted phenyl group (246TMP-3SF5) has significant potential as anticancer drug as the pentafluorosulfanyl group has been shown to improve compound stability and biological activity (Altomonte und Zanda 2012; Goehringer, Peng, u. a. 2021). In HCC cell lines 246TMP-3SF5 showed pronounced antiproliferative effects with low micromolecular IC₅₀ values after 48 hours. Long-term assessment of the antiproliferative efficacy also revealed powerful effects on long-term proliferation with close to no colonies formed after growth for 14 days in both HepG2 and HuH-7 cells. Moreover, unspecific cytotoxic effects could be excluded as predominant mode of action (Viperino u. a. 2025).

In this work, we further investigated the underlying mechanism of action of 246TMP-3SF5 in HCC cell lines HepG2 and HuH-7. Additionally, experiments were performed to examine the effects of the new drug on cell cycle regulation and crucial hallmarks of cancer cells, such as cell migration and induction of angiogenesis. Furthermore, we evaluated the antineoplastic efficacy of the novel inhibitor on HCC microtumors in-ovo to assess its effectiveness as potential novel drug in the treatment of HCC.

2. Materials and Methods

2.1. Compounds

The synthesis of the pentafluorosulfanyl-substituted chalcone 246TMP-3SF5 was published before (Viperino u. a. 2025). The HSP90 inhibitor 17-AAG was purchased from AbMole BioScience (Houston, USA, Cat. No. M2320) and the multikinase inhibitor sorafenib was purchased from TargetMol (Boston, USA, Cat. No. T0093L). Inducer of ferroptotic cell death erastin was purchased from TargetMol (Boston, USA, Cat. No. 571203-78-6) and the ferroptosis inhibitor ferrostatin-1 was purchased from Bio-Techne GmbH (Wiesbaden-Nordenstadt, Germany, Cat. No. 5180). 20 mM stock solutions of the compounds were prepared using dimethyl sulfoxide (DMSO, Sigma-Aldrich, Darmstadt, Germany) and stored at -20 °C. Working solutions (1 mM) were freshly prepared for the experiments by diluting the stock solution with cell culture medium and stored at 4 °C for up to two weeks.

2.2. Cell Culture

Human hepatocellular carcinoma cell line HuH-7 (JCRB number: JCRB0403) and human hepatoma cell line HepG2 (ATCC number: HB-8065) were cultured in Roswell Park Memorial Institute (RPMI) 1640 Medium (Gibco, Thermo Fisher Scientific, Waltham, USA) supplemented with 10% fetal bovine serum, 2 mM L-glutamine (Corning, Arizona, USA), 100 U/mL penicillin and 100 µg/mL streptomycin (Bio&SELL, Feucht, Germany) at 37 °C, 5% CO₂ and a humidified atmosphere in an incubator.

2.3. Antiproliferative assessment by cristal violet assay

Treatment-induced antiproliferative effects on HCC cell lines were measured as a decrease in cell count assessed by crystal violet (hexamethyl-para-rosaniline chloride from Sigma-Aldrich, Darmstadt, Germany) staining as previously described (Goehringer, Peng, u. a. 2021; Ma u. a. 2022). In brief, HepG2 and HuH-7 cells were seeded in 96-well plates and allowed to grow for 72 hours in an incubator (37 °C, 5% CO2, humidified atmosphere). Thereafter, cells were treated with rising concentrations of 246TMP-3SF5, 17-AAG or sorafenib. After 24, 48 and 72 hours the cells were fixed using 1% glutaraldehyde and stained for 30 minutes with 0.1% crystal violet. Excess unbound dye was removed by rinsing with water for 30 minutes. Cells were then lysed by incubation with 0.2% Triton X-100 (Sigma-Aldrich, Darmstadt, Germany) overnight, dissolving the bound dye. The extinction of crystal violet was measured at 570 nm using ELISA-readers (Dynex Technologies, Denkendorf, Germany; GloMax^® Discover Microplate Reader Promega, Madison, USA). Non-linear regression analysis was performed and dose-response curves generated. Calculated IC₅₀ values are given as means ± SD. A minimum of n > 3 independent experiments performed in triplicates or more for each cell line.

2.4. Caspase-3 activity assessment

To assess drug-induced apoptosis caspase-3 assay was performed using Ac-DEVD-AMC (Cayman Chemical, Michigan, USA, Cat. No. 14986), a substrate which exhibits a fluorogenic signal upon cleavage by active caspase-3. The assay was performed as previously described (Goehringer, Peng, u. a. 2021). Briefly, cells were seeded at a density of 1 × 10⁵ cells/well in 12-well plates and incubated overnight. Afterwards, the cells were treated with 1 µM, 5 µM, 10 µM of 246TMP-3SF5 and 17-AAG for 24 hours. Cells were washed and frozen, followed by a lysis with RIPA buffer for 30 minutes on ice. Using the PierceTM BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, USA, Cat. No. 23227) the protein content of the samples was quantified. To detect the activity of caspase-3 the samples were incubated at 37 °C for 1 hour with Ac-DEVD-AMC. Substrate cleavage was measured fluorometrically using the GloMax^® Discover Microplate Reader (Promega, Madison, USA). Measured fluorescent signal was then normalized to the previously calculated protein levels in each sample. n = 3–4 independent experiments were performed in duplicates, and data are given as folds of the control ± SEM.

2.5. Western Blot Analysis

Changes in the level of protein expression were assessed through western blotting. HepG2 and HuH-7 cells were seeded at a density of 1x10⁶ cells/well in 100 mm petri dishes and incubated until reaching desired confluency of 70%. Drugs were added at given concentrations for 24 hours. Untreated cells served as control. After washing with PBS, the cells were frozen at -20 °C. Subsequently, radioimmunoprecipitation assay (RIPA) buffer mixed with cOmplete™ Mini Protease Inhibitor (1 tablet/10 mL) (Roche Diagnostics, Mannheim, Germany, Cat. No. 11836170001) was added to lyse the cells. Following these steps, the protein levels were analyzed using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, USA, Cat. No. 23227) and protein levels were normalized to ensure equal protein loading of 20 µg/lane. Laemmli buffer and β-mercaptoethanol were added to the probes following denaturation at 96 °C for 10 minutes. Proteins were separated using SDS–PAGE and electro-transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were blocked in 5% non-fat dry milk in Tris-buffered saline with 0.1% Tween-20 and incubated with primary antibody overnight at 4 °C. Primary antibodies include poly (ADP-ribose) polymerase (Cell Signaling Technology, Boston, MA, USA, Cat. No. 9532, 1:2000) and β-actin (Sigma-Aldrich, Munich, Germany, Cat. No. A5441, 1:8000). Secondary peroxidase-coupled anti-IgG anti-mouse (926-80010 LI-COR Biotechnology GmbH, Bad Homburg, Germany, 1:10000) or anti-rabbit (926-80011 LI-COR Biotechnology GmbH, Bad Homburg, Germany, 1:10000) antibodies were incubated for 1 hour at room temperature. Bands were visualized using Clarity Western Blotting ECL Substrates (Bio-Rad Laboratories, Munich, Germany) and ChemiDoc MP Imaging System (Bio-Rad Laboratories, Munich, Germany). Blots of n = 3 independent experiments were generated, and representative images are shown.

2.6. Cell cycle analysis by flow cytometry

To determine changes in the cell cycle due to the HSP90 inhibitors cell cycle analysis was performed using the DNA-staining dye propidium iodide (PI) and performing flow cytometry. HepG2 and HuH-7 cells were seeded at 5 × 10³ cells/well in 6-well plates and incubated overnight and then treated with 246TMP-3SF5 and 17-AAG at the indicated concentrations for 24 hours. Cells were washed with PBS and harvested, fixed and permeabilized with 70% ethanol overnight at -20 °C. Thereafter, cells were washed three times with PBS and incubated with RNase A (0,4 mg/mL) for 30 minutes at 37 °C. After staining the cells with PI (0,5 µg/mL) (Invitrogen, Eugene, OR, USA) in the dark, flow cytometry was conducted using the FACSCanto II (BD Biosciences, Heidelberg, Germany). 50000 single cells were measured and analyzed using FlowJo software (version 10.10.0; LLC, Ashland, OR, USA). n = 4 independent experiments were performed, representative histograms and data shown as means ± SEM is shown.

2.7. Reactive oxygen species detection

To assess the formation of cytosolic reactive oxygen species (ROS) the cell membrane-permeable dye CellROX^® Orange (Thermo Fisher Scientific, Waltham, USA, Cat. No. C10443) was used. The dye exhibits a strong red fluorescent signal upon oxidation at excitation/emission levels of 545 nm/565 nm. Cells were seeded (1 × 10⁵ cells/well for HepG2, 2 × 10⁵ cells/well for HuH-7) in 12-well plates and allowed to attach overnight. The cells were then treated with 246TMP-3SF5, 17-AAG or erastin at the indicated concentrations. CellROX^® Orange (1 μM) was applied together with the drugs. Following an incubation time of 24 hours fluorescent images were taken using a ZOE™ Fluorescent Cell Imager (Bio-Rad Laboratories, Munich, Germany). Untreated cells served as negative control, while untreated cells incubated with 1.6 mM H₂O₂ for 30 minutes prior to imaging served as positive control. n > 3 independent experiments were performed, and representative images are shown.

2.8. Measurement of intracellular glutathione levels

To determine changes of the intracellular glutathione levels GSH-Glo™ Glutathione Assay was performed (Promega, Madison, USA, Cat. No. V6911) according to the manufacturer’s instructions. In brief, HCC cells were seeded at 3 × 10⁶ cells/well and allowed to adhere overnight before being treated with 246TMP-3SF5, 17-AAG or erastin at the indicated concentrations for 24 hours. Intracellular glutathione drives the formation of luciferin, which then produces light directly proportional to the amount of glutathione present. Luminescence was measured using the GloMax^® Discover Microplate Reader (Promega, Madison, USA). n = 3–4 independent experiments were performed in duplicates and data are given as the mean percentage compared to control (set to 100%) ± SEM.

2.9. Malondialdehyde Analysis

To assess lipid peroxidation Malondialdehyde (MDA) Colorimetric Assay Kit was purchased from Elabscience (Houston, USA, Cat. No. E-BC-K028-M). The assay was used according to the manufacturer’s instructions. In brief, cells treated with 246TMP-3SF5, 17-AAG, erastin and ferrostatin-1 at the indicated concentrations for 24 hours. Afterwards cells were harvested, homogenized and incubated with thiobarbituric acid at 100 °C for 40 minutes. The thiobarbituric acid reacts with MDA to produce a pinkish coloring which can be measured at absorption of 532 nm. Protein levels of samples were measured using Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, USA, Cat. No. 23227). Measured absorbance was normalized to the protein level of the samples, and the MDA concentrations were calculated using the provided standard solution. Erastin (10 µM) served as positive control. n = 3–5 independent experiments were performed, and data are shown as means ± SEM compared to the control (set to 100%).

2.10. Scratch Assay

To investigate anti-migratory effects of 246TMP-3SF5 HepG2 and HuH-7 cells were seeded in 6-well plates and allowed to grow until sub-confluency. The cell monolayer was scratched vertically with a 10 µL pipette tip and detached floating cells were removed by gently washing with PBS. Cells were then incubated for up to 24 hours with rising concentrations of 246TMP-3SF5 (1-10 µM) or 10 µM 17-AAG, respectively. Untreated cells served as control. t = 0 hours and t = 24 hours pictures were taken using the EVOS M5000 microscope (Thermo Fischer Scientific, Waltham, USA). Scratch area was quantified using the wound healing size tool plug-in for ImageJ. Cell migration was calculated using the following formula

$$\begin{array}{l}CellMigrationRate\left[\%\right]\\ =\left({\displaystyle \frac{initialscratcharea\left[\%\right]-scratchareaafter24hours\left[\%\right]}{initialscratcharea\left[\%\right]}}\right)\times 100\%.\end{array}$$

Quantified cell migration rate in percentage as means ± SEM and representative images of n = 3 independent experiments are shown.

2.11. In-ovo assessment of antiangiogenic and antineoplastic effects

To examine anti-angiogenic and anti-neoplastic effects of 246TMP-3SF5 in a systemic setting in-ovo experiments utilizing the chorioallantoic membrane (CAM) of fertilized chicken eggs were conducted as previously described (Goehringer, Peng, u. a. 2021; Ma u. a. 2022; Goehringer, Biersack, u. a. 2021). Fertilized chicken eggs (Gallus gallus) were incubated at 37.8 °C and >58% humidity. On day 3–4 of embryonic development a small hole was pierced into the poles of the egg and 1–2 mL of albumin were removed causing the CAM to detach from the eggshell and lower itself down into the allantoic cavity.

To assess anti-angiogenic effects changes in the vascularization of the chorioallantoic membrane (CAM) upon treatment were analyzed. On day 11 of embryonic development the eggshell was carefully opened to place a sterilized silicone ring (Ø 5 mm) on the region of interest on the CAM. Afterwards, 20 µL 0.9% NaCl, which served as control, or compound diluted in 0.9% NaCl was added into the ring at the indicated concentrations. Changes in the angiogenesis and vascularization of the CAM were documented by taking pictures using a light microscope equipped with a Kappa digital camera system (Distelkamp-Electronic, Kaiserslautern, Germany) after 24 and 48 hours. Representative images are shown. At least n = 5 eggs for each condition were assessed.

To assess anti-neoplastic effects of the novel compound 3 × 10⁶ HuH-7 tumor cells suspended in 20 µL Matrigel were implanted on the chorioallantoic membrane of the fertilized chicken eggs on day 8 of embryonic development using a silicone ring of 5 mm in diameter. The tumor bearing eggs were incubated for 24 hours allowing the tumors to attach and connect to the microvessels of the CAM, following a topical application of 20 µL of 246TMP-3SF5 or 17-AAG for 72 hours at the indicated concentrations. Viability of the chicken embryo and tumor growth were controlled daily. Pictures of the tumors were taken using a microscope (Kappa digital camera system, Distelkamp-Electronic, Kaiserslautern, Germany). Thereafter, the tumors were carefully cut out and weighed. Representative images are shown and tumor weight after treatment is quantified as means ± SEM. A minimum of n = 5 eggs were analyzed for each condition.

2.12. Molecular docking

The docking studies of 17-AAG and 246TMP-3SF5 bound to the ATP binding site of HSP90 (PDB ID 1YET) was performed to determine its binding mode (Stebbins u. a. 1997). Protein and ligand preparation and visualization was done using UCSF Chimera (Pettersen u. a. 2004). For docking AutoDock Vina was used (Trott und Olson 2010). Crystal structure from PDB ID 1YET with 1.90Å resolution was extracted and opened in UCSF Chimera. All the water molecules were removed and protein and ligand molecule were separately prepared. Preparation of protein was carried out by adding Hydrogens, missing side-chains and Gasteiger charges (J. Wang u. a. 2006). All the ligand molecules are prepared, converted to PDBQT files and docked using AutoDock Vina.

2.13. Molecular dynamics simulation

We performed 5ns MD simulation of 246TMP3SF5 docked into ATP-binding site of HSP90 protein using Gromacs (Lemkul 2024). The system was parameterized using Amber forcefield and solvated in 10Å cubic box with TIP3 model. Counter ions added to neutralize the system. Energy minimization done using steepest decent method. Thermostats were kept at 300K with NVT and 1 atm pressure for NPT ensemble runs done for 100 ps each and the production run done for 5 ns. Simulation was forwarded to production run after the temperature and pressure equilibration was attained.

2.14. Statistical Analysis

Visualization of data and statistical analyses were performed using GraphPad Prism 10 (GraphPad Software Version 10.6.1, San Diego, CA USA). All data are expressed as mean ± SEM, unless stated otherwise. Ordinary one-way ANOVA followed by Dunnett’s or Tukey’s multiple comparisons post-hoc test were performed for statistical analyses. Statistical significance was defined as * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001.

3. Results

3.1. Antiproliferative effects of 246TMP-3SF5 against human liver cancer cells

It was previously reported that the novel compound 246TMP-3SF5 exerts selective antiproliferative effects on liver cancer cells with low micromolecular IC50 values of 1.3 µM (± 0.2 µM) in HepG2 and 2.2 µM (± 0.3 µM) in HuH-7 cells after 48 h of treatment (Viperino u. a. 2025). In comparison, established HSP90-inhibitor 17-AAG amounted to similar IC₅₀ values of 1.6 µM (± 0.5 µM) and 0.4 µM (± 0.3 µM) for HepG2 and HuH-7 cells respectively. However, sorafenib, currently used as chemotherapeutic agent in first-line treatment of advanced-stage HCC, reached IC₅₀ values of 4.4 µM (± 0.9 µM) in HepG2 cells and 5.4 µM (± 0.6 µM) in HuH-7 cells.

The antiproliferative effect of 246TMP-3SF5 was time- and dose-dependent in both HCC cell lines (Figure 1). It is noteworthy that increased treatment time enhanced the antiproliferative effects, however the differences could only be observed at concentrations starting from 1.0 µM and are modest at 5.0 and 10.0 µM after 48 and 72 hours. The dose-dependently increased antiproliferative effectiveness is more pronounced, following a sigmoid curve. In HepG2 cells the effect of the novel inhibitor is more prominent, 1.0 µM 246TMP-3SF5 inhibiting proliferation down to 50% compared to control after 72 hours. Interestingly, the antiproliferative efficacy reached a peak at 5.0 µM 246TMP-3SF5 at 72 hours. After 72 hours HCC cell proliferation upon treatment amounted to only 1.5% (± 0.3%) of HepG2 cells and 5.9% (± 1.8%) of HuH-7 cells. While a twofold increase in concentration to 10.0 µM 246TMP-3SF5 did increase the antiproliferative effect further to almost no viable cells left, the effect at 5.0 µM is already profound.

In line with the antiproliferative effects observed after 24 hours of treatment, concentrations of 1.0, 5.0 and 10.0 µM 246TMP-3SF5 were used for the following experiments regarding the investigation of the mode of action.

3.2. Apoptotic effects of 246TMP-3SF5 in HCC cell lines

To investigate the mode of action of the novel HSP-inhibitor 246TMP-3SF5 in HCC cell lines apoptotic cell death was evaluated as possible mechanism of action explaining the antiproliferative effects of the novel inhibitor. Investigation of apoptosis was conducted via assessment of subG1 peak via flow cytometry, caspase-3 activity assay as well as western blot assessment of poly-(ADP-ribose)-polymerase (PARP). Dose-dependent effects of 246TMP-3SF5 were evaluated throughout the experiments. 17-AAG (tanespimycin) was used as reference HSP90 inhibitor for comparison.

246TMP-3SF5 exhibited a significant dose-dependent increase of cells in subG1 peak in both HCC cell lines after treatment for 24 hours, the effect more pronounced than that for 17-AAG at equal concentration (Figure 2 A, B). SubG1 peak reached a mean of up to 27.7% (± 2.3%) in HepG2 and 26.3% (± 1.2%) in HuH-7 cells upon treatment with 10.0 µM 246TMP-3SF5. These findings suggest that 246TMP-3SF5 increased the number of dead cells and DNA fragments upon treatment. Therefore, apoptosis was further investigated as potential mode of action by analyzing key enzymes of the apoptosis cascade. We found caspase-3, a key effector caspase during apoptosis, to be significantly elevated after treatment with the novel inhibitor 246TMP-3SF5 for 24 hours (Figure 2 C, D). The effect showed a dose-dependent pattern for HepG2 cells, with both HCC cell lines reaching a peak of caspase-3 activity of more than threefold compared to untreated controls at 10.0 µM 246TMP-3SF5. While at high concentrations 17-AAG also slightly increased caspase-3 activity by up to 1.6-fold (± 0.04) in HepG2 and 1.7-fold (± 0.1) in HuH-7 cells compared to control, the effect following treatment with 246TMP-3SF5 was significantly more pronounced. In comparison, the novel inhibitor increased caspase-3 activity by up to 3.7-fold (± 0.3) in HepG2 and 3.8-fold (± 0.2) in HuH-7 cells at equal concentration of 10.0 µM. Western blot analysis revealed a decrease of full-length PARP and cleavage of PARP following treatment with 10.0 µM 246TMP-3SF5 for 24 hours in both HCC cell lines. 17-AAG also induced the cleavage of PARP in HepG2 cells, said effect only faintly noticeable in HuH-7 cells (Figure 2 E,F).

3.3. Assessment of cellular stress and ferroptotic cell death in HCC cell lines treated with 246TMP-3SF5

To further decipher the underlying molecular mode of action of the antiproliferative effects of 246TMP-3SF5 ferroptotic cell death was assessed. Ferroptosis is a form of regulated cell death characterized by an iron-dependent accumulation of lipid peroxides and subsequent plasma membrane damage of cells, driven by an accumulation of reactive oxygen species (ROS) and a dysregulation of antioxidative systems, such as glutathione – glutathione peroxidase 4 pathway (Tang und Kroemer 2020; Jiang u. a. 2021; Dixon und Olzmann 2024). The ferroptosis-inducer erastin was used as positive control.–

Firstly, ROS induction by 246-TMP-3SF5 in HCC cells was investigated (Figure 3). Upon treatment with 246TMP-3SF5 for 24 hours, an increase of fluorescent red labeled cells due to a rise of intracellular ROS levels compared to control became evident. ROS induction by 246TMP-3SF5 was more pronounced in HuH-7- than in HepG2 cells (Figure 3 B). Erastin, showed a similar effect as 246TMP-3SF5, its impact on ROS accumulation clearly noticeable in HuH-7 cells as well. In contrast, 17-AAG induced a less intense ROS accumulation in HuH-7 cells. These findings indicate that 246TMP-3SF5 increased intracellular ROS levels predominantly in HuH-7 cells, similar to the ferroptosis-inducer erastin.

Glutathione (GSH) levels and malondialdehyde (MDA) concentrations of HCC cells treated with the novel inhibitor were analyzed to further evaluate ferroptosis as relevant mode of action. GSH is a cysteine-containing tripeptide playing an essential role as electron donor in antioxidative systems of the cell suppressing ferroptosis. Low GSH levels cause am imbalance in antioxidative systems, impairing their function and leading to ferroptosis (Jiang u. a. 2021; Dixon und Olzmann 2024; X.-D. Zhang u. a. 2023). Erastin induces ferroptosis by blocking the main channel responsible for cellular cysteine import, which, in turn, then serves as a building block for GSH (Jiang u. a. 2021). Moreover, reactive aldehydes formed through lipid peroxidation such as MDA are crucial indicators of ferroptosis (Tang u. a. 2021).

Assessment of intracellular GSH levels revealed a significant dose-dependent GSH depletion in both HCC cell lines following treatment with 246TMP-3SF5 (Figure 4 A, B). Interestingly, the results showed 1.0 µM 246TMP-3SF5 increased GSH levels compared to control. Strikingly, 10.0 µM 246TMP-3SF5 significantly decreased intracellular GSH levels down to 3.6% (± 0.4%) in HepG2 and 9.6% (± 0.5%) in HuH-7 cells compared to control. Similar effects were observed for 10.0 µM erastin, GSH levels of 5.7% (± 0.6%) for HepG2 and 0.4 (± 0.1%) for HuH-7 cells. On the other hand, 10.0 µM 17-AAG also caused a reduction in GSH levels, however the effect was not as pronounced with GSH levels of 72.8% (± 4.8%) and 61.9% (± 3.2%) in HepG2 and HuH-7 cells respectively. Comparing both inhibitors, 246TMP-3SF5 significantly exceeded ferroptosis-associated GSH level reduction in HCC cells (Figure 4 A, B). The results indicate that oxidative stress and imbalance of antioxidative systems might also play a role in the mechanisms behind the antiproliferative effect of 246TMP-3SF5 in HCC cell lines.

Furthermore, changes in MDA levels were analyzed. Upon treatment with 246TMP-3SF5 a significant dose-dependent increase of MDA levels was observed in both cell lines, suggesting that ferroptosis is a relevant form of cell death occurring due to 246TMP-3SF5 treatment (Figure 4 C, D). MDA levels reached a mean of 233.6% (± 15.0%) in HepG2 and 221.9% (± 14.0%) in HuH-7 cells at 10.0 µM 246TMP-3SF5 compared to untreated controls. Notably, the increase of MDA due to 17-AAG turned out to be minor at 105.0% (± 17.7%) in HepG2 and 122.9% (± 1.3%) in HuH-7 cells, in line with the previously assessed GSH depletion not as pronounced. It is noteworthy that while both erastin and the novel inhibitor led to a significant increase in MDA levels in both HCC cell lines at 10.0 µM, the effects is slightly more pronounced for 246TMP-3SF5. Thereafter reversibility of the observed effect was assessed by co-incubation with ferroptosis-inhibitor ferrostatin-1. Increase of MDA levels was partially reversible for erastin and 246TMP-3SF5 in both HCC cell lines upon co-incubation with ferrostatin-1 (Figure 4 E, F). The effect was significant in HuH-7 cells for the novel inhibitor.

3.4. Evaluation of cell cycle regulation upon 246TMP-3SF5 treatment in HCC cell lines

Cell cycle arresting effects of 246TMP-3SF5 were determined by flow cytometry. Treating HCC cell lines with 246TMP-3SF5 revealed a disruption of the cell cycle compared to control (Figure 5). With increasing concentrations of the novel inhibitor G0/G1- and G2/M-phase increased. At 1.0 µM 246TMP-3SF5 S-phase was augmented in HepG2 cells (Figure 5 A, C). Upon treatment of HuH-7 cells with rising concentrations of 246TMP-3SF5 changes in the cell cycle phases were observed, with the G2/M-phase increased compared to control (Figure 5 B, D). 17-AAG led to a distinct G0/G1-phase arrest in both HCC cell lines, while 246TMP-3SF5 did not cause a clear cell cycle arrest in a specific phase but rather a complex disruption of the cell cycle.

3.5. 246TMP-3SF5 inhibits cell migration in HCC cell lines

Migration of cells is a hallmark of cancer cells essential for local invasion of tissue and to form metastases, thus a novel promising compound inhibiting tumor cell migration would be of particular importance (Hanahan und Weinberg 2011). Anti-migratory effects of 246TMP-3SF5 were investigated through scratch assay in HCC cell lines, pictures taken at t = 0 h and t = 24 hours after the initial scratch and cell migration rates were assessed through the area of the initial gap filled by migrated HCC cells after 24 hours.

Following treatment of HCC cell lines with the novel inhibitor 246TMP-3SF5 a dose-dependent decrease in tumor cell migration was observed (Figure 6). In HepG2 cells 10.0 µM 246TMP-3SF5 significantly decreased cell migration resulting in a migration rate of only 2.2% (± 1.1%). The effect was found to be approximately equally as pronounced as the antimigratory effect of 10.0 µM 17-AAG at a 2.5% (± 1.2%) migration rate (Figure 6, A, C). An even more prominent antimigratory effect of 246TMP-3SF5 was observed in HuH-7 cells with a decrease in cell migration rate by up to 18.4% (Figure 6 B, D). Our findings suggest that the novel compound 246TMP-3SF5 impairs tumor cell migration in HCC cell lines, significantly decreasing cancer cell migration rate compared to control.

3.6. Anti-angiogenic effects of 246TMP-3SF5 in-ovo

Inducing angiogenesis allows tumors to access higher amounts of nutrients and oxygen sustaining continuous tumor mass growth (Hanahan und Weinberg 2011). Changes in angiogenesis and vascularization of the chorioallantoic membrane (CAM), a highly vascularized membrane in fertilized chicken eggs, were evaluated upon topical application of 246TMP-3SF5.

Topical treatment with 10.0 µM 246TMP-3SF5 for up to 48 hours caused morphological irregularities of blood vessels of the CAM. Blood vessels collapsed decreasing in diameter (Figure 7, blue arrows) and vessel branch points appeared as blunt ends of nodule-like structure (Figure 7, black arrows). Overall anti-angiogenic effects of 246TMP-3SF5 became apparent, the vessels of the CAM treated with the novel inhibitor showing a notable reduction of newly formed micro-vessels and prominent irregularities in the blood vessel network compared to control. Similar effects were observed for 17-AAG as well. In addition, even at the highest concentration of 10.0 µM of 246TMP-3SF5 no toxicity was observed in chicken embryos of the embryonic development days from 11 to 13.

3.7. Anti-neoplastic efficacy of 246TMP-3SF5 in-ovo

Following previous assessments of the antiproliferative efficacy of 246TMP-3SF5 in HCC cell lines resulting in very promising results as novel anticancer drug (Viperino u. a. 2025), anti-neoplastic efficacy of the novel inhibitor in-ovo on fertilized chicken eggs was evaluated. HCC cell line HuH-7 derived microtumors were grown on the CAM of fertilized chicken eggs and treated with 246TMP-3SF5 or 17-AAG. Tumor growth was observed over 72 hours and final tumor weight quantified. 246TMP-3SF5 treatment significantly decreased tumor growth and reduced tumor weight in-ovo in a dose-dependent manner compared to untreated controls (Figure 8 A, B). Tumor weight dropped to a mean (± SEM) of 10.6 mg (± 1.5 mg) at 5.0 µM and 7.4 mg (± 1.2 mg) at 10.0 µM 246TMP-3SF5 respectively. The results showed a decrease of tumor growth as well as reduction of tumor weight for 17-AAG as well. However, at 10.0 µM 17-AAG mean tumor weight amounted to 13.7 mg (± 1.9 mg), a similar antineoplastic efficacy observed for 246TMP-3SF5 at only 5.0 µM. Therefore the novel inhibitor exeeded the anti-neoplastic efficacy of 17-AAG in-ovo (Figure 8 B).

3.8. Molecular docking of 246TMP-3SF5 into Heat Shock Protein 90

HSP90 is a homodimer molecular chaperone in physiological state, with each subunit containing N-terminal domain (NTD) that contains the ATP/ADP binding site which is crucial for the chaperone’s activity, Middle domain (MD) that is involved in interactions with client proteins and co-chaperones and C-terminal domain (CTD) that is responsible for dimerization, forming the interface between the two HSP90 subunits (Stebbins u. a. 1997) (Figure 9 A). It undergoes ATP-driven conformational changes to regulate client protein folding and stability. The docking protocol was first validated by redocking the co-crystallized ligand geldanamycin (GDM), which reproduced key interactions with Asn51, Lys58, Asp93, Lys112, Gly135, and Phe138, yielding a heavy-atom RMSD of 0.202 Å relative to its crystal conformation. As expected, 17-AAG, a semi-synthetic derivative of geldanamycin, adopted a similar binding orientation within the ATP-binding pocket of HSP90 and formed hydrogen-bond interactions with Asp54, Lys58, Asp93, Lys112, and Phe138, with a binding affinity of −9.6 kcal/mol (Figure 9B). The conserved binding mode of GDM and 17-AAG validated the docking protocol and provided a structural basis for evaluating the binding of the newly designed compound 246TMP-3SF5.

Docking of 246TMP-3SF5 revealed a hydrogen-bond interaction with Lys58 and a binding affinity of −7.3 kcal/mol (Figure 9C). Surface visualization showed that the phenyl–SF5 moiety was deeply accommodated within a hydrophobic pocket lined by residues such as Met98, Leu107, Val150, and Val186, contributing additional stabilization through favorable hydrophobic (van der Waals) interactions (Figure 9D). To assess the influence of the pentafluorosulfanyl group on binding, the nitro analogue SU086 was also docked into the HSP90 ATP-binding site. SU086 exhibited a comparable binding affinity of −7.3 kcal/mol and formed a hydrogen bond with Phe138; however, it adopted a distinct binding orientation relative to 246TMP-3SF5 (Supplementary Figure SI). Surface analysis indicated that the phenyl–NO2 group of SU086 was oriented toward the more polar region of the binding pocket near Asp54 and Ser53, whereas the phenyl–SF5 group of 246TMP-3SF5 occupied a predominantly hydrophobic region. The altered binding mode of SU086 is likely attributable to the rigid chalcone framework combined with the different steric and electronic properties of the nitro substituent.

3.9. Molecular dynamics simulation of 246TMP-3SF5 into Heat Shock Protein 90

The structural stability of the HSP90–246TMP-3SF5 complex was assessed by monitoring the root mean square deviation (RMSD) of the protein backbone and ligand throughout the molecular dynamics simulation. The backbone RMSD remained stable at approximately 1.5 Å, indicating that the overall protein structure was well maintained during the simulation (Figure 10 A). Similarly, the ligand RMSD stabilized around 2.2 Å, suggesting that 246TMP-3SF5 retained a consistent binding orientation within the ATP-binding pocket of HSP90 (Figure 10 B). Final frame also shows the H-bond with Lys58 and -SF5 group properly occupying the pocket formed by Leu48, Asn51, Leu107, Lys112, Phe138, Val150 and Val86 (Figure 10 C). Following an initial equilibration phase, both protein and ligand RMSD values exhibited only minor fluctuations, demonstrating the attainment of a stable conformational state. The relatively low RMSD values and absence of significant deviations indicate that the protein–ligand complex remained structurally stable throughout the simulation, with no evidence of ligand dissociation or major conformational rearrangements. The observed ligand fluctuations are likely attributable to local adjustments within the binding pocket that facilitate optimal protein–ligand interactions while preserving the overall binding mode.

4. Discussion

HSP90 is a promising therapeutic target in HCC treatment as it stabilizes various client proteins involved in hepatocarcinogenesis and is often overexpressed, correlates with advanced tumor stage and poor prognosis (X. Liu u. a. 2016; Cheng u. a. 2015). Antiproliferative assessment of the novel pentafluorosulfanyl-substituted chalcone 246TMP-3SF5, derived from the HSP90 inhibitor SU086, revealed pronounced time- and dose-dependent antiproliferative effects against HCC cell lines with low micromolecular IC₅₀ values at 1.3 (± 0.2 µM) for HepG2 and 2.2 (± 0.3 µM) for HuH-7 cells. Interestingly, previous assessments revealed the mother drug SU086 to not be similarly effective in HCC cell models with IC₅₀ values above 10.0 µM (Viperino u. a. 2025). The data suggests the modification to the original chalcone structure positively impacted the biological antiproliferative activity of 246TMP-3SF5 against HepG2 and HuH-7 cells. In comparison 17-AAG achieved similar IC₅₀ values in HepG2 cells, while being more effective in HuH-7 cells. It is noteworthy that the novel inhibitor reached lower IC₅₀ values compared to sorafenib, which is currently used in first-line treatment of advanced-stage HCC.

The investigation of the mode of action of 246TMP-3SF5 revealed a multitude of effects in liver cancer cell models HepG2 and HuH-7. Investigating apoptosis a significant dose-dependent increase of cells in subG1 peak, significant elevation of apoptosis key caspase caspase-3 activity and cleavage of PARP was observed, indicating apoptosis to be a relevant mode of cell death caused by treatment of HCC cells with 246TMP-3SF5. It is noteworthy that although the HSP90 reference inhibitor 17-AAG also showed an increase in subG1 peak and elevation of caspase-3 activity, the effect at equal concentrations of 10.0 µM compared to 246TMP-3SF5 was not as prominent. Preclinical studies have shown 17-AAG to induce apoptosis, however the described pathways differ from the ones assessed in our study. 17-AAG seems to induce apoptosis through degradation of multiple HSP90 client proteins, mainly via the intrinsic mitochondrial pathway, although differences were reported depending on the cell line which was assessed (Nimmanapalli u. a. 2003, 17-; Mitchell u. a. 2007).

Additionally, our findings revealed increased intracellular ROS levels, significant GSH depletion, as well as significant dose-dependent MDA concentration elevation suggesting ferroptosis occurred upon treatment with 246TMP-3SF5. Although 10.0 µM 17-AAG also reduced intracellular GSH levels, the effect is significantly stronger for 246TMP-3SF5 at equal concentration. Notably, 17-AAG causes almost no increase in MDA concentration upon treatment. The observed GSH depletion and rise of MDA concentration due to treatment with 246TMP-3SF5 and ferroptosis-inducer erastin were similarly pronounced. This hints at a difference regarding 246TMP-3SF5’s and 17-AAG’s ability to induce ferroptosis. 17-AAG has been shown to react chemically with GSH forming adducts depleting intracellular GSH, in line with our findings (Cysyk u. a. 2006). Although GSH is an essential cofactor to GPX4, part of the Xc⁻/GSH/GPX4 system, to date no study demonstrated 17-AAG directly induced ferroptosis in HCC cells, consistent with our findings. On the other hand, our results indicate 246TMP-3SF5 induces ferroptosis in HCC cell lines, however further experiments are needed to reveal further details.

Further investigations revealed a 246TMP-3SF5 induced cell cycle disruption, although no clear cell cycle arrest was observed. In HuH-7 cells an increase of the G2/M-phase upon treatment with 246TMP-3SF5 was noticeable. In contrast, 17-AAG caused HCC cell arrest in the G0/G1–phase of the cell cycle. Other studies also found 17-AAG to cause cell cycle arrest. Interestingly, an increase of HuH-7 cells in G0/G1–phase up to 55.9% (± 7.1%) alongside a G2/M–phase arrest with 28.9% (± 13.0%) was observed,explained by the authors through cdc2 degradation (Watanabe u. a. 2009).

HCCs have been shown to be highly vascularized tumors with HSP90 overexpression promoting angiogenesis, thus a drug exhibiting antiangiogenic effects is of particular interest (X. Liu u. a. 2016; Cheng u. a. 2015; Meng u. a. 2017). Our results show that 246TMP-3SF5 exerts significant anti-migratory as well as anti-angiogenic effects in-ovo. Moreover, molecular docking into HSP90 demonstrated the phenyl-SF5 moiety stably accommodated within a hydrophobic pocket of the ATP-binding site, revealing the likely form of interaction of 246TMP-3SF5 with HSP90. Inhibition of HSP90 leads to proteasomal degradation of its client proteins, many of which are of high importance for angiogenesis and migration of tumor cells. HSP90 inhibitors impair multiple pro-angiogenic pathways, such as reducing the expression of VEGF receptors on endothelial cells. Additionally, crucial for angiogenesis, HIF-1α is also part of the long list of HSP90 client proteins. Amongst others it drives VEGF transcription under hypoxic conditions, allowing for tumor angiogenesis. Through HIF-1α degradation, inhibition of HSP90 leads to impaired endothelial recruitment and angiogenesis, inhibiting a crucial pro-angiogenic pathway of tumor cells (Sanderson u. a. 2006; Isaacs u. a. 2002; P.-C. Zhang u. a. 2020; Staufer und Stoeltzing, o. J.) . Similarly, HSP90 inhibitors impair tumor cell migration by causing proteasomal degradation of important receptor tyrosine kinases, such as EphA2, which is involved in cellular motility and invasiveness (Shinji Tsutsumi und Neckers 2007; S Tsutsumi u. a. 2008; Song u. a. 2010; Nagaraju u. a. 2015; Annamalai u. a. 2009). Additionally, a fraction of HSP90α is secreted into the extracellular space playing an important role in cell migration, cytoskeletal reorganization as well as extracellular matrix remodeling.

Moreover, 246TMP-3SF5 also significantly reduced HCC microtumor growth in-ovo at 5.0 and 10.0 µM by approximately 50% compared to control, showing great potential as novel drug for HCC treatment.

Many different HSP90 inhibitors have been developed and preclinically as well as clinically evaluated against various cancer entities. Some examples evaluated in HCC cell models include geldanamycin derivates 17-AAG, 17-DMAG, as well as resorcinol-based inhibitor AUY922, non-quinone inhibitor PU-H71, SNX-2112 and STA-9090 (Cheng u. a. 2015; Breinig u. a. 2009; X. Wang u. a. 2014; L. Liu u. a. 2021; Leng u. a. 2012). Often preclinical evaluations reveal promising effects against various cancers; however clinical translation remains highly challenged. Thus far pimitespib is the only HSP90 inhibitor approved for clinical treatment (Hoy 2022). It received official approval for the treatment of gastrointestinal stromal tumor patients refractory to tyrosine kinase inhibitor therapy in Japan in 2022 (Rastogi u. a. 2024; Goel u. a. 2025; Chang u. a. 2024). To this day no HSP90 inhibitor has been approved for treatment of HCC. Main challenges in clinical trials are dose-limiting toxicities, hepatotoxicity most relevant in HCC treatment where liver function is often already impaired. For geldanamycin derivates such as 17-AAG, the benzoquinone moiety revealed itself to be central for the hepatotoxicity observed. Patients in clinical trials showed dose-limiting hepatotoxicity through aminotransferase elevation (Samuni u. a. 2010; Ramanathan u. a., o. J.; Solit u. a., o. J.). In contrast 246TMP-3SF5 is derived from the HSP90 inhibitor SU086, containing a chalcone backbone and previous investigations showed no increased toxicity in HCC cell models, even at high concentrations of the drug (Viperino u. a. 2025). Representing an in-vivo setting our in-ovo experiments showed no increased mortality of chicken embryos, suggesting 246TMP-3SF5 exhibits no remarkable toxicity, however further in-vivo experiments evaluating toxicity and specifically hepatotoxicity are needed.

Regarding the clinical application of HSP90 inhibitors, combining HSP90 inhibitors with targeted therapies, immune checkpoint inhibitors or chemotherapy has shown synergistic effects improving anticancer activity. Moreover, combination therapy allows for the decrease of the dose of each individual drug therefore also trying to reduce toxicity. Various combinations are currently being investigated in clinical trials, with some reporting improved clinical response to treatment (Rastogi u. a. 2024; Kim u. a. 2025). Another interesting strategy being investigated is combining HSP90 inhibitor therapy with heat stress through thermal therapy. Previously, studies have shown combining hyperthermia with other treatment modalities to be beneficial (Veltsista u. a. 2023). Preclinical evaluation showed HSP90 inhibitor STA-9090 to sensitize HCC to hyperthermia-induced DNA damage (L. Liu u. a. 2021). Therefore, investigating how the combination of 246TMP-3SF5 with other drugs used in HCC treatment or hyperthermia can impact the anticancer activity is of great interest.

5. Conclusions

In summary, our findings in this study provide important insight into the mechanisms of action of the promising novel inhibitor 246TMP-3SF5. Our results indicate apoptosis and ferroptosis play an important role in the antiproliferative effects observed. Following treatment of HCC cell lines with 246TMP-3SF5 cell cycle disruption was observed and the hallmarks of cancer cell migration and induction of angiogenesis were significantly impaired. Lastly, our results indicate a significant-treatment induced reduction of microtumor growth in-ovo, overall proving 246TMP-3SF5 to be a promising novel compound meriting further investigations and evaluation as new anticancer drug in the treatment of hepatocellular carcinoma.