Panc02 cells (murine ductal pancreatic adenocarcinoma) were grown in RPMI 1640 medium supplemented with 10% heat inactivated fetal bovine serum (FBS), 2 mM glutamine, 1% sodium pyruvate and 1% penicillin/streptomycin (Gibco, Life Technologies). Cells were incubated overnight at a density of 60,000 cells per well in 12-well plates at 37°C in 95% air/5% CO₂ until confluency was reached. Hereafter, cells were treated with UAMC-2526 (10 µM) in the absence or presence of gemcitabine (10 µM) for 32 hours. For measuring the autophagy flux 160 nM of bafilomycin A₁ (Santa Cruz Biotechnology, sc-2021550) was added for 2 hours.

Panc02 cells were incubated overnight at a density of 60,000 cells per well in 12-well plates after which UAMC-2526 was added at different concentrations in the presence or absence of 10 µM gemcitabine. Cell proliferation was measured using neutral red solution as previously described (35). Briefly, cells were incubated for 2 hours with 0.1% neutral red solution at 37°C in 95% air/5% CO₂. Subsequently, cells were washed with PBS and neutral red was extracted by addition of 0.05 M NaH₂PO₄ in 50% ethanol. After 3 minutes, optical density was read at 540 nm using a microtiter plate reader.

To detect cell death, 1 µg/ml propidium iodide (PI, Molecular Probes) and 10 µg/ml Hoechst (Life Technologies) was added, followed by visualization of PI/Hoechst-labeled cells using a Celena S digital Imaging System (Logos Biosystems).

4×10⁶ viable Panc02 cells (suspended in 100μl PBS) were injected into the right hind leg of female C57/BL6 mice (n= 40, 6-7 weeks old; Charles River Laboratories). All animals received a numerical code throughout the whole experiment and were kept under environmentally controlled conditions (12 hours light/dark cycle, 20–24°C and 40–70% relative humidity) with food and water ad libitum. When tumors reached an approximate volume of 150 mm³, mice were randomly divided into 4 groups (10 mice per group) for treatment with clinical-grade gemcitabine (Selleck Chemicals), UAMC-2526, gemcitabine combined with UAMC-2526, or vehicle (DMSO/propylene glycol/ethanol 50:40:10). Gemcitabine (60 mg.kg^-1) was administered i.p. 3x/week for 28 days. Unidose solutions of compound UAMC-2526 were prepared in vehicle according to the individual mouse weight and loaded into snap-fit-sealed polypropylene microvials. Solutions were prepared the day before treatment and transferred to osmotic minipumps (model 1002, Alzet) for 2-week applications achieving an in vivo pumping rate of 3 mg.kg^-1.day^-1. The control and gemcitabine group received vehicle administered via a 100 µl osmotic minipump system. The osmotic minipumps were replaced after 2 weeks to obtain a total treatment period of 28 days. Mice not treated with gemcitabine received an i.p. injection with saline at the same time.

Body weight was measured three times a week and tumor growth was monitored three times a week with digital caliper measurements starting upon detection of palpable tumors. Tumor volume was calculated using the modified ellipsoid formula [i.e., ½ (length×width²)]. Relative tumor volume (RTV) was calculated as V_x/V₁, with V_x, volume at time point t, and V₁, the volume at baseline. The percentage of tumor growth inhibition (TGI) was calculated as 1- (RTVx/RTVc)x100 with RTVx being the relative tumor volume from the treated group and RTVc the relative tumor volume from the control group. The tumor doubling time was calculated as V= V₀2^t/d. Tumor samples were isolated on day 28 or when tumors reached a volume of 1500 mm³. The study was conducted in accordance with the Declaration of Helsinki and was approved by the ethics committee of the University of Antwerp (file number 2017-68).

¹⁸F-FDG was produced in an automated module of the Department of Radiopharmacy at Antwerp University Hospital using a FASTlab FDG Cassette (GE Healthcare). Small animal PET imaging was performed using two Siemens Inveon PET-CT scanners (Siemens Preclinical Solution). The acquired PET data were reconstructed using 4 iterations with 16 subsets of the 2D ordered subset expectation maximization (OSEM 2D) algorithm following Fourier rebinning (FORE). Normalization, dead time, random CT-based attenuation and single-scatter stimulation (SSS) scatter corrections were applied. CT voltage and amperage were set to 80 keV and 500 μA, respectively. Animals received two PET scans: one baseline (before start of the treatment) and one follow-up PET scan, 6-7 days after start of the treatment. Mice first received an IV bolus injection of ¹⁸F-FDG (37 MBq) in the left lateral tail vein, immediately followed by a double measurement (Easy-Touch, Lifescan, France) of pre-scan whole blood glucose levels, from a drop of blood collected from the contralateral tail vein. The average value was considered to correct for blood glucose levels. Twenty-five minutes after the ¹⁸F-FDG injection (i.e. uptake period), mice were anesthetized with isoflurane (5% for induction and 2% for maintenance) and five minutes after the start of the anesthesia mice were prepared and positioned onto the PET scanner, resulting in a total uptake period of 30 min. Sequentially to the PET scan, a 7 min CT scan was performed for attenuation as well as scatter correction purposes and for anatomical localization of the tumors. To minimize variability, all mice were handled under the same anesthetic agent, incubation time, uptake period and body temperature. Mice were not fasted as fasting can severely induce autophagy. Several quantification methods were applied to quantify ¹⁸F-FDG uptake in the tumors. Regions of interest (ROIs) were drawn on the CT images manually by qualitative assessment covering the whole tumor slice-by-slice. Tumor volume and tracer uptake were generated by summation of voxels within the tomographic planes using PMOD Software 6.3 (PMOD technologies) and expressed by mean and max standardized uptake values (SUVmean or SUVmax). The standardized uptake value (SUV) was thereby defined as (CT×W)/Dinj, where CT is radioactivity in the tissue (kBq/cc), W is weight of the animal (g) and Dinj is injected dose (kBq). SUVmean represents the mean concentration of the tracer and was calculated with and without correction for the glucose (Glc) level: (CT×W×Glc)/Dinj. SUV max represents the concentration in the 5 hottest pixels. In addition, total lesion glycolysis (TLG) was calculated by multiplying the SUVmean and the tumor metabolic volume (TMV). A threshold method was also applied to quantify tracer tumor uptake on the PET image, whereby an elliptic volume-of-interest that enclosed the entire tumor was positioned manually and was centered on the tumor area that showed maximal radiotracer uptake. Then 3-dimensional isocontours at 60% of the maximum pixel value within this volume-of-interest were drawn automatically.

Tissue samples were fixed in 4% neutral buffered formalin (Sigma-Aldrich) for 24 hours prior to paraffin embedding. To assess tumor cell proliferation, tumor sections (5 µm thick) were incubated with primary antibodies against MCM2 (ab108935). Besides immunohistochemistry, tumor sections were also stained with hematoxylin-eosin and acellular areas histologically analyzed to quantify the percentage of tumor necrosis. Apoptosis was determined via TUNEL assay using the ApopTag Plus Peroxidase In situ Apoptosis kit (Merck, S7101). All images were acquired with an Olympus U-TUIX-2 microscope and analyzed with ImageJ 1.42 software (National Institutes of Health). Analyses were carried out without prior knowledge of the treatments.

Tumor samples were prepared for TEM analysis as previously described (36). A FEI Tecnai microscope was used to examine ultrathin sections at 80-120 kV. Autophagic vacuoles were quantified on 40 different images taken at random per sample.

Tumor samples were mixed in RIPA lysis buffer (Sigma-Aldrich, R0278) containing protease and phosphatase inhibitors (Roche Diagnostics, 046931590001; 049606845001) using a Percellys 24 tissue homogenizer (Bertin Instrument). After centrifugation, the protein concentration of the supernatant of each sample was determined using the bicinchoninic acid method (BCA, ThermoFischer). Laemmli buffer (Bio-Rad, 1610737) containing 5% β-mercaptoethanol (Sigma-Aldrich) was added and samples were heat-denatured for 5 minutes in boiling water. Cell culture samples were directly lysed in Laemmli sample buffer (Bio-Rad, 1610737) containing 5% β-mercaptoethanol (Sigma-Aldrich) and heat-denatured for 5 min in boiling water. Next, equal amounts of protein were separated on pre-cast Novex Bolt 12% Bis-Tris gels (Invitrogen, NW00125BOX) and blotted onto Immobilion-FL polyvinylidene fluoride membranes. After incubation in Odyssey^® Blocking Buffer (Li-Cor Biosciences, LI 927-50000), membranes were probed overnight at 4°C with primary antibodies diluted in Odyssey Blocking Buffer, followed by a one hour incubation with IRDye-labeled secondary antibodies at room temperature. Antibody detection was achieved using an Odyssey SA infrared imaging system (LI-COR Biosciences). The intensity of the protein bands was quantified using Emperia software. Primary antibodies included anti-p62/SQSTM1 (Abcam, ab56416), anti-LC3 (Nanotools, 0231-100/LC3-5F10) and anti-ACTB (Abcam, ab115777). IRDye-labeled secondary antibodies (goat anti-mouse [IgG926-68070] and goat anti-rabbit [IgG926-32211]) were purchased from LI-COR Biosciences.

Results are expressed as mean ± SEM. All statistical analyses were performed using SPSS software (version 27, SPSS Inc). Statistical tests are specified in the figure legends. Differences were considered significant at p ≤ 0.05.

To investigate the potential beneficial effects of UAMC-2526, Panc02 cells were treated in vitro with UAMC-2626 in the presence or absence of gemcitabine. Cell proliferation was inhibited by UAMC-2526 in a concentration and time dependent way ( Figure 1A ). Combined treatment of Panc02 cells with UAMC-2526 and gemcitabine almost completely prevented cell proliferation. To determine whether the inhibition of cell proliferation was related to cell death induction, propidium iodide (PI) labeling experiments were performed ( Figure 1B ). UAMC-2526 did not induce cell death in a concentration range between 0.1-10 µM. However, administration of UAMC-2526 at a higher concentration (50 µM) combined with longer incubation times (48 and 72 hours) showed a significant increase in PI positive cells, even though the number of PI positive cells remained limited (<25%). Gemcitabine (10 µM) initiated cell death after 72 hours of treatment. Combined treatment of Panc02 cells with UAMC-2526 and gemcitabine showed an additive, but no synergistic effect on cell death induction.

Western blots were performed on cell lysates of Panc02 cells to analyze the expression of autophagy marker LC3-II. Administration of UAMC-2526 decreased basal autophagy in Panc02 cells as shown by a decreased conversion of the soluble isoform LC3-I into the autophagosome-associated isoform LC3-II. Gemcitabine treatment initiated autophagosome formation, as demonstrated by elevated levels of LC3-II, but this effect was significantly blunted in the presence of UAMC-2526 ( Figure 2 ). Moreover, both UAMC-2526 and gemcitabine stimulated p62 accumulation ( Figure 2 ). Inhibition of autophagic flux by adding bafilomycin A1 blocked LC3-II degradation and allowed the accumulation of LC3-II in untreated cells. However, bafilomycin A1 did not further stimulate LC3-II levels in gemcitabine-treated cells ( Figure 3 ).

The effect of UAMC-2526 on tumor growth was further explored in vivo using C57/BL6 mice inoculated with Panc02 cells. When tumors reached an approximate volume of 150 mm³ (D0), mice were randomly divided into four treatment groups: vehicle, gemcitabine, UAMC-2526 and UAMC-2526 combined with gemcitabine. Body weight did not differ during the treatment between groups ( Figure 4A ), albeit relative tumor volume (RTV) rapidly increased in the vehicle and UAMC-2526 treated groups from D14 ( Figure 4B ). Compared to vehicle treated controls, mice treated with gemcitabine showed a reduced RTV from D19. Combined administration of gemcitabine and UAMC-2526 further reduced RTV compared to vehicle and monotreatment with gemcitabine from D14.

Next to RTV, the percentage tumor growth inhibition (%TGI) was calculated ( Table 1 ). Administration of gemcitabine led to an average %TGI of 27.4 ± 4.0 between D24-D31, while combined treatment with UAMC-2526 resulted in a significantly higher average %TGI of 44.5 ± 1.9 (***p ≤ 0.001, Dunnett’s T3 Multiple Comparison Test, n = 39-40). In addition, tumor doubling time increased significantly after combined treatment with gemcitabine and UAMC-2526 ( Figure 5 ).

According to PET/CT imaging, all treatment groups showed a clear ¹⁸F-FDG uptake with no significant differences between groups at baseline (i.e. before treatment; Figure 6A ). After 6-7 days of treatment, no significant differences in ¹⁸F-FDG uptake between treatment groups could be observed (irrespective of the PET quantification method, Figure 6B ). None of the quantification methods showed significant results. Also a correction for blood glucose levels did not change significances. Representative PET/CT images are shown in Figure 6C .

Microscopic analysis of gemcitabine-treated tumor samples revealed a modest, albeit significant reduction in immunoreactivity for cell proliferation marker MCM2 as compared to vehicle-treated controls ( Figure 7 ). Also tumors of mice treated with UAMC-2526 and gemcitabine showed a significant decrease in immunoreactivity for MCM2 as compared to the control group. However, there was no significant difference in MCM2 immunostaining in tumors of UAMC-2526 treated mice. The frequency of tumor apoptosis between treated animals and controls did not change ( Figure 8 ), though there was a significant reduction of necrosis in tumors of mice treated with UAMC-2526 ( Figure 9 ). Western blots were performed on lysates of tumor samples to analyze the autophagosomal marker protein LC3-II and selective autophagy receptor p62 ( Figure 10 ). Neither LC3-II nor p62 levels were significantly different between groups. Similarly, electron microscopy did not show any significant differences in the formation of autophagic vacuoles in tumor tissues of the combination treatment group as compared to the controls ( Figure 11 ).