Polyethyleneglycol-Betulinic Acid (PEG-BA) Polymer-Drug Conjugate Induces Apoptosis and Antioxidation in a Biological Model of Pancreatic Cancer

Pancreatic cancer (PC) is one of the most aggressive solid malignancies with poor treatment response and low survival rates. Herbal medicines such as betulinic acid (BA) have shown potential in treating various solid tumours, but with limitations that can be circumvented by polymer-drug conjugation. Polyethylene glycol-BA (PEG-BA) polymer-drug conjugate has previously shown selective anticancer activity against PC cells. Here, we elucidate the mechanism of cell death and the cell death pathway, anti-inflammatory and antioxidant activities of PEG-BA. PEG-BA induced apoptotic cell death by arresting MIA-PaCa-2 cells in the Sub-G1 phase of the cell cycle compared with BA and untreated cells (39.50 ± 5.32% > 19.63 ± 4.49% > 4.57 ± 0.82%). NFκB/p65 protein expression was moderately increased by PEG-BA (2.70 vs. 3.09 ± 0.42 ng/mL; p = 0.1521). However, significant (p < 0.05) overexpression of the proapoptotic genes TNF (23.72 ± 1.03) and CASPASE 3 (12,059.98 ± 1.74) compared with untreated cells was notable. The antioxidant potential of PEG-BA was greater (IC50 = 15.59 ± 0.64 µM) compared with ascorbic acid (25.58 ± 0.44 µM) and BA-only (>100 µM) and further confirmed with the improved reduction of hydroperoxide levels compared with BA-only (518.80 ± 25.53 µM vs. 542.43 ± 9.70 µM). In conclusion, PEG-BA activated both the intrinsic and extrinsic pathways of apoptosis and improved antioxidant activities in PC cells, suggesting enhanced anticancer activity upon conjugation.


Introduction
Pancreatic cancer (PC) is a lethal solid malignancy with increasing incidence and mortality worldwide [1]. Most patients present with advanced stages of the disease, in which case the tumour is inoperable and has metastasised to other sites or both [2]. Currently, the best curative option is surgery using adjuvant chemotherapy, with only 15% of patients presenting with resectable or early-stage disease [2]. Combination chemotherapy regimens, including gemcitabine and FOLFIRINOX (oxaliplatin, irinotecan, fluorouracil and leucovorin), are used as the first-line therapy for PC and have shown improved patient survival [3]. Although survival has improved, this is still insufficient as the survival rate is still abysmal compared with other cancers. Furthermore, studies have demonstrated that the administration of chemotherapy could lead to adverse side effects suggesting the need to develop potent drugs against PC but with reduced side effects, including toxicity to normal cells [4].

Cell Cycle Status Determination with Propidium Iodide Dye
The single parameter-based propidium iodide staining technique was used to assess the cell cycle status of the cells, pre-and post-treatment with vehicle control (0.33% DMSO), 4 µM of BA and PEG-BA and 4.5 µM of gemcitabine following 72 h of treatment (37 • C; 95% humidity, 5% CO 2 ). The cells were harvested using trypsin EDTA and washed (200× g). The cell pellet was resuspended with ice-cold 1X PBS followed by dropwise addition of icecold 70% ethanol for fixation at −20 • C for at least 24 h before staining. A total of 500 µL of the staining mixture, prepared by dissolving 2 mg of RNase A in 0.1% Triton-X100 reagent in PBS and containing 400 µL of 500 µg/mL propidium iodide dye, was added to each sample and incubated at room temperature for 30 min. Data acquisition commenced at low flow rates (< 300 events/sec) for improved resolution and adequate DNA quantification on the BD LSRFortessa™ Analyser and FACSDiva software (BD Biosciences, Franklin Lakes, NJ, USA). At least 30,000 events were recorded per sample. The fluorescent signal from the samples was detected using the 610/20 band filter pass filter. The generated Flow Cytometry Standard (FCS) files in FACSDiva were imported and analysed using FlowJo v10.8.1 (FlowJo, LLC, Ashland, KY, USA).

Gene Expression Analysis of Apoptotic Genes 2.7.1. Extraction of Total RNA
Total RNA was extracted using the E.Z.N.A. ® Total RNA Kit I (Omega Bio-Tek, Norcross, GA, USA) according to the manufacturer's guidelines. Briefly, 700 µL of guanidium thiocyanate (GTC)-lysis buffer containing 2% β-mercaptoethanol was added to the pelleted cells previously treated with 4 µM BA and PEG-BA and 4.5 µM gemcitabine for 72 h. This mixture was then transferred to an RNA homogeniser mini-column and centrifuged for 60 s at 13,000× g. This was followed by adding 700 µL of freshly prepared 70% ethanol and transferring to a HiBind ® RNA mini-column. The mixture was centrifuged at 10,000× g for 60 s until the RNA in-solution was bound to the column. The bound RNA was washed (30 s at 10,000× g) with 500 µL of RNA Wash Buffer I. The filtrate was discarded, and the RNA was rewashed (60 s at 10,000× g) using a second buffer (RNA Wash Buffer II containing 100% ethanol). This step was repeated once to ensure proper washing of the RNA before drying the HiBind ® RNA mini-column to remove any residual ethanol (15,000× g for 120 s). The HiBind ® RNA mini-column was transferred to a 1.5 mL Eppendorf tube, and 60 µL of nuclease-free water was added, followed by centrifugation (15,000× g for 120 s) to elute the RNA. The eluted RNA was assessed for purity, and the concentration was determined using the Nanodrop 2000 (Thermo Fischer Scientific, Waltham, MA, USA).

Complementary DNA (cDNA) Synthesis
Total RNA (250 ng/µL) was used to make the cDNA template for each sample. The cDNA synthesis step was performed using the ProtoScript II First Strand cDNA Synthesis Kit (New England BioLabs Inc, Ipswich, MA, USA) following the manufacturer's protocol. Briefly, 20 µL of the reaction mixture was added to respective Eppendorf tubes: 2 µL of d(T)23 VN (50 µM) primers, 10 µL of the ProtoScript II reaction mix (2X), 2 µL of the Protoscript II enzyme mix (10 X) and 6 µL of the RNA template diluted with the required amount of nuclease-free water to have a uniform final cDNA concentration for all the samples of 1000 ng/µL. The samples were then incubated in a SimpliAmp Thermal Cycler (Thermo Fischer Scientific, Waltham, MA, USA) under the following conditions: 42 • C for one hour, enzyme inactivation at 80 • C for 5 min, and an infinite hold at 4 • C.

Quantitative Real-Time Polymerase Chain Reaction Analysis
Real-time PCR was performed in a 96-well plate format (on ice) according to the MIQE guidelines [43]. Briefly, 5 µL of the PowerUp™ SYBR™ Green Master Mix (2X),  Table 1), 2 µL of the cDNA template, and 2 µL of nuclease-free water were added to each well of a 96-well PCR plate, in duplicate. The universal PCR conditions used entailed a UDG activation at 50 • C for 2 min, Dual-lock DNA polymerase activation at 95 • C for 2 min, PCR cycling (40 cycles) of denaturation (95 • C for 15 s), annealing primer-specific temperature range of 50-60 • C for 15 s and extension (72 • C for 60 s). Two reference genes: β-Actin (ACTB) and mitochondrial ribosomal protein L19 (MRPL19), were used to normalise relative fold expression. Fold changes were determined using the 2 −∆∆CT method [44]. Table 1. List of genes used for real-time polymerase chain reaction.

Gene Name
Forward Primer Sequence Reverse Primer Sequence The housekeeping genes used to normalise fold expression.

NF-κB/p65 Transcription Factor Assay
The human nuclear factor kappa-B p65 subunit (NF-κB/p65) enzyme-linked immunosorbent assay (ELISA) kit from Elabscience Biotechnology Inc. (Houston, TX, USA) was used to determine the concentration of NF-kB/p65 present in cellular samples after treatment with complete DMEM, 4 µM of BA and PEG-BA, and 4.5 µM gemcitabine (72 h incubation). Cell supernatants and lysates derived from these samples were used to perform the ELISA according to the manufacturer's protocol with slight modifications. Briefly, 100 µL of the standards (0-10 ng/mL), lysates (diluted 2X with sample diluent), and supernatants were added to respective wells in the protein-coated 96-well plate, in duplicate, sealed, and incubated at 37 • C for 1.5 h. The samples were discarded, and 100 µL of the biotinylated detection antibody working solution was added and incubated for an hour. A wash step was employed by adding 350 µL of the wash buffer, soaking for at least 60 s, and aspirating three times. Horseradish peroxidase (HRP) conjugate working solution (100 µL) was added to each well, sealed, and incubated for 30 min. The plate was washed five times, and 90 µL of the substrate reagent was added to the wells. The plates were sealed and covered in foil to protect them from light. The plate was incubated for at least 15 min, followed by adding 50 µL of the stop solution. The optical density was measured at 450 nm using a Multiscan Sky microplate reader (Thermo Fischer Scientific, Waltham, MA, USA).

ROS Reduction and Potential Antioxidant Analysis
Although the aetiology of PC is not well-known, inflammation has been described as an important risk factor [2,45]. Two assays were used to assess the potential antioxidative capacities of BA and PEG-BA. This was performed using the 2,2-Diphenyl-1picrylhydrazyl (DPPH) antioxidant assay for the compound activity and the N,N-Dimethylp-phenylenediamine (DEPPD) assay for ROS activity in MIA PaCa-2 cells.
2.9.1. DPPH Assay BA, PEG-BA, and the ascorbic acid standard were tested in the antioxidant activity assay at concentrations in the range 0.78-100 µM. This assay was modified from the 2014 protocol by Mistry and Shah [46]. Briefly, 180 µL of distilled water was added to the first nine rows of a 96 well-plate and 100 µL to the remaining wells. This was followed by adding 20 µL of BA, PEG-BA and ascorbic acid (dissolved in methanol) in the appropriate wells in a fume hood to prevent evaporation of the methanol. A two times serial dilution was performed before adding 100 µL of freshly prepared methanolic DPPH solution to triplicate wells, except for the compound control wells, in the dark. The plates were covered in foil and shaken on a Titertek microplate shaker (Medical EXPO, Marseilles, France) for 10 min. This was followed by an additional 35 min incubation period in the dark. After the incubation, absorbance readings were obtained at 517 nm using a Multiscan Sky microplate reader (Thermo Fisher Scientific, Waltham, MA, USA).

DEPPD Assay
The DEPPD assay was used to determine reactive oxygen metabolites (hydroperoxide) produced by reactive oxygen species (ROS) in PC cells. The oxidative status and ROS reduction potential of the MIA PaCa-2 cells pre-and post-treatment (72 h) with 0.33% DMSO and optimised working concentrations: 8 µM for BA and PEG-BA and 9 µM gemcitabine, were determined using the DEPPD assay. The assay was performed according to the protocol by William (2012), with modifications to accommodate using cell supernatants instead of the standard urine or blood samples [47]. Briefly, 140 µL of 0.1 M sodium acetate buffer (pH 4.8) was added to a 96-well plate. This was followed by adding 10 µL of 0.39-50 µM hydrogen peroxide standard and cell supernatants of negative control of (media only), vehicle control, BA and PEG-BA and standard gemcitabine. One hundred microlitres of R 1 /R 2 solution (100 µg/mL DEPPD (R 1 ) and 4.37 µM FeSO4 (R 2 ) at a ratio of 1:25) was added to all wells except the blank. The plate was covered with foil for at least 1 min and analysed using a Victor™ X3 model 2030 multilabel plate reader (PerkinElmer, Waltham, MA, USA) at a wavelength of 505 nm with a kinetic loop of 30 readings per well.

Statistical Analysis
The GraphPad Prism Version 8 software (GraphPad, San Diego, CA, USA) was used to determine the statistical significance of the data. To assess whether the data were normally distributed, normality tests such as the Shapiro-Wilk (W), Anderson-Darling (A2*), D'Agostino-Pearson omnibus (K2), and Kolmogorov-Smirnov (distance) that are available in the GraphPad Prism (version 8) software were used. It was assumed that a p-value greater than α (0.05) meant that the null hypothesis (H 0 ) was accepted, and the data were classified as following a normal distribution. Furthermore, quantile-quantile (Q-Q) plots were assessed with the assumption that plotting predicted residual vs. actual residual forms a roughly straight line for normally distributed data. For two groups (passed normality tests and Q-Q plots), an unpaired Student t-test was performed. A two-way analysis of variance (ANOVA) was performed for more than two groups, followed by multiple comparisons using the Tukey post hoc test, if the data passed the normality tests and Q-Q plots. For data that only passed the normality tests but not the Q-Q plots, the non-parametric Kruskal-Wallis was performed followed by Dunn's post hoc test. A p-value of less than 0.05 was considered statistically significant.

PEG-BA Causes Cell Rounding in a Dose and Time-Dependent Manner
At least 100,000 cells/mL were visualised for morphological changes before and after treatment using light microscopy. Figure 1A-I show representative micrographs of the effect of BA and PEG-BA on MIA PaCa-2 cells at 72 h only. Quantitative graphs showing percentage cytotoxicity using MTT are shown in Figure 1J. The results showed that, compared with untreated cells ( Figure 1A), 2 and 4 µM BA started to induce minimal cell rounding or clumps. However, most plated cells were still viable even at 72 h (4.88 ± 2.42% and 5.55 ± 3.85% cytotoxicity- Figure 1C,D). Cell rounding is associated with apoptosis due to the initial detachment from neighbouring cells and shrinkage [48]. At the highest tested concentration (8 µM), there was a slightly increased cytotoxic effect (15.70 ± 5.70%) that was not visible in the photomicrographs as the cells continued to grow ( Figure 1E). On the other hand, the conjugated compound (PEG-BA) resulted in the formation of cell clumps at the lowest tested concentration (1 µM- Figure 1F) and a cytotoxic effect higher than that induced by BA at the same concentration (p = 0.0255). Therefore, PEG-BA induced increased cell clumping and cytotoxicity in a concentration and time-dependent manner. At 72 h, 8 µM of PEG-BA resulted in clumping of the cells forming clusters, indicating high toxicity towards the PC cells. This was also observed at 24 and 48 h ( Figures S1 and S2). This is further confirmed with a 100% cytotoxicity after 72 h. effect of BA and PEG-BA on MIA PaCa-2 cells at 72 h only. Quantitative graphs showing percentage cytotoxicity using MTT are shown in Figure 1J. The results showed that, compared with untreated cells ( Figure 1A), 2 and 4 µM BA started to induce minimal cell rounding or clumps. However, most plated cells were still viable even at 72 h (4.88 ± 2.42% and 5.55 ± 3.85% cytotoxicity- Figure 1C,D). Cell rounding is associated with apoptosis due to the initial detachment from neighbouring cells and shrinkage [48]. At the highest tested concentration (8 µM), there was a slightly increased cytotoxic effect (15.70 ± 5.70%) that was not visible in the photomicrographs as the cells continued to grow ( Figure 1E). On the other hand, the conjugated compound (PEG-BA) resulted in the formation of cell clumps at the lowest tested concentration (1 µM- Figure 1F) and a cytotoxic effect higher than that induced by BA at the same concentration (p = 0.0255). Therefore, PEG-BA induced increased cell clumping and cytotoxicity in a concentration and time-dependent manner. At 72 h, 8 µM of PEG-BA resulted in clumping of the cells forming clusters, indicating high toxicity towards the PC cells. This was also observed at 24 and 48 h ( Figures  S1 and S2). This is further confirmed with a 100% cytotoxicity after 72 h.

PEG-BA Induces a Sub-G 1 Arrest in Pancreatic Cancer Cells
The cell cycle status of the MIA PaCa-2 cells before and after treatment was performed using propidium iodide dye on 4 µM of BA and PEG-BA treated cells. This is a concentration in the IC 50 range from this ( Figure S3) and a previous study [8]. The untreated MIA PaCa-2 cells showed a high percentage of cells in the G 1 /G 0 (45.27 ± 4.70%) compared with cells undergoing DNA synthesis (S: 17.67 ± 1.22%) and those preparing to enter mitotic division or already in the phases of mitosis (G 2 /M: 25.07 ± 2.44%) (Figure 2). The remaining cells, which were a minority (4.57 ± 0.82%), were in the area termed the sub-G 1 phase, which is the phase to the left of the G 1 /G 0 phase. These results served as baseline populations for comparison after treatment, where a reduction or increase in cells from one phase to another was regarded as due to treatment [49]. At 4 µM, BA shifted the MIA PaCa-2 cells into sub-G 1 (19.63

PEG-BA Induces a Sub-G1 Arrest in Pancreatic Cancer Cells
The cell cycle status of the MIA PaCa-2 cells before and after treatment was performed using propidium iodide dye on 4 µM of BA and PEG-BA treated cells. This is a concentration in the IC50 range from this ( Figure S3) and a previous study [8]. The untreated MIA PaCa-2 cells showed a high percentage of cells in the G1/G0 (45.27 ± 4.70%) compared with cells undergoing DNA synthesis (S: 17.67 ± 1.22%) and those preparing to enter mitotic division or already in the phases of mitosis (G2/M: 25.07 ± 2.44%) (Figure 2). The remaining cells, which were a minority (4.57 ± 0.82%), were in the area termed the sub-G1 phase, which is the phase to the left of the G1/G0 phase. These results served as baseline populations for comparison after treatment, where a reduction or increase in cells from one phase to another was regarded as due to treatment [49].

PEG-BA Elevates the Expression of Proapoptotic Genes
The cell cycle data reported here and the apoptosis assay performed previously [8] suggest that the anticancer activity of PEG-BA is via apoptosis. We then further investigated the potential mechanism by assessing the expression of key apoptotic genes (TNF, TNFSF10, CASPASES (2,3,7,8), BID and BAX) of the extrinsic and intrinsic pathways of apoptosis using quantitative real-time PCR. Compared with BA, PEG-BA treatment in PC cells resulted in a significant increase in the proapoptotic genes, CASPASE 2 (0. 58   G1), purple (G1/G0), yellow (S) and green (G2/M) peaks. Double-sided arrows were inserted to highlight the sub-G1 and S phases. Overall, both BA and PEG-BA resulted in a shift of cells from the G1/G0 to the sub-G1 phase, indicating apoptosis induction. (D) Quantitative plots representing the overall cycling status of the PC cells (mean ± SEM), n = 3, **: p < 0.01. The Inkscape software (Version 1.1) was used to edit figure 2A-C for clarity.

PEG-BA Elevates the Expression of Proapoptotic Genes
The cell cycle data reported here and the apoptosis assay performed previously [8] suggest that the anticancer activity of PEG-BA is via apoptosis. We then further investigated the potential mechanism by assessing the expression of key apoptotic genes (TNF, TNFSF10, CASPASES (2,3,7,8), BID and BAX) of the extrinsic and intrinsic pathways of apoptosis using quantitative real-time PCR. Compared with BA, PEG-BA treatment in PC cells resulted in a significant increase in the proapoptotic genes, CASPASE 2 (0. 58

PEG-BA Treatment May Result in NF-κB Expression in Pancreatic Cancer Cells
The protein expression levels of nuclear factor kappa B (p65-subunit) in both lysate and supernatant of treated (PEG-BA, BA only and gemcitabine) MIA PaCa-2 cells were quantified using a sandwich ELISA. A higher level of NF-κB was observed in the lysate of the untreated and treated MIA PaCa-2 cells compared with the corresponding supernatant ( Figure 4). NF-κB levels remained unchanged in the supernatant (untreated: 1.12 ± 0.06 ng/mL; gemcitabine: 1.09 ± 0.03 ng/mL; BA: 1.13 ± 0.06 ng/mL and PEG-BA: 1.17 ± 0.07 ng/mL), whereas the lysate showed moderate increases with the treatment of PEG-BA (3.09 ± 0.42 ng/mL; p = 0.1521) and BA (2.78 ± 0.27 ng/mL; p = 0.9981) compared with the untreated sample (2.70 ± 0.21 ng/mL). Gemcitabine resulted in a slight but insignificant decrease (2.67 ± 0.19 ng/mL; p = 0.9484).

PEG-BA Treatment May Result in NF-κB Expression in Pancreatic Cancer Cells
The protein expression levels of nuclear factor kappa B (p65-subunit) in both and supernatant of treated (PEG-BA, BA only and gemcitabine) MIA PaCa-2 cell quantified using a sandwich ELISA. A higher level of NF-κB was observed in the of the untreated and treated MIA PaCa-2 cells compared with the corresponding natant ( Figure 4). NF-κB levels remained unchanged in the supernatant (untreated 0.06 ng/mL; gemcitabine: 1.09 ± 0.03 ng/mL; BA: 1.13 ± 0.06 ng/mL and PEG-BA: 0.07 ng/mL), whereas the lysate showed moderate increases with the treatment o BA (3.09 ± 0.42 ng/mL; p = 0.1521) and BA (2.78 ± 0.27 ng/mL; p = 0.9981) compare the untreated sample (2.70 ± 0.21 ng/mL). Gemcitabine resulted in a slight but insign decrease (2.67 ± 0.19 ng/mL; p = 0.9484).
The level of ROS, specifically hydroperoxides, in the MIA PaCa-2 cells before and after treatment with PEG-BA, BA only and gemcitabine was assessed using the DEPPD assay. Overall, the results showed that there was a significantly (p < 0.0001) higher concentration of hydroperoxides (7.32-fold) in untreated cancer cells compared with the normal cells and treated cells (Figure 6). Treating the MIA PaCa-2 cells with PEG-BA resulted in a moderate reduction in hydroperoxide levels compared with the untreated (Figure 6)  The level of ROS, specifically hydroperoxides, in the MIA PaCa-2 cells before and after treatment with PEG-BA, BA only and gemcitabine was assessed using the DEPPD assay. Overall, the results showed that there was a significantly (p < 0.0001) higher concentration of hydroperoxides (7.32-fold) in untreated cancer cells compared with the normal cells and treated cells ( Figure 6). Treating the MIA PaCa-2 cells with PEG-BA resulted in a moderate reduction in hydroperoxide levels compared with the untreated (Figure 6), with no statistical significance (p = 0.5066). Notably, gemcitabine treatment reduced hydroperoxide levels (433.34 ± 33.53 µM) compared with untreated (546.96 ± 29.03 µM; p = 0.0451), PEG-BA treated (518.80 ± 25.53 µM; p = 0.2011) and BA-only (542.43 ± 9.70 µM; p = 0.0581) treated cancer cells.   The level of ROS, specifically hydroperoxides, in the MIA PaCa-2 cells before and after treatment with PEG-BA, BA only and gemcitabine was assessed using the DEPPD assay. Overall, the results showed that there was a significantly (p < 0.0001) higher concentration of hydroperoxides (7.32-fold) in untreated cancer cells compared with the normal cells and treated cells ( Figure 6). Treating the MIA PaCa-2 cells with PEG-BA resulted in a moderate reduction in hydroperoxide levels compared with the untreated (Figure 6

Discussion
Pancreatic cancer is associated with chemoresistance. The conjugation of PEG to BA has been shown to kill PC cells effectively and selectively and could circumvent chemoresistance [8]. Analysing critical cellular processes such as apoptosis, inflammation and antioxidation during treatment with PEG-BA can help delineate the molecular mechanisms associated with treatment with the conjugated drug. In this study, PC cells were treated with PEG-BA and its proapoptotic and antioxidative characteristics were demonstrated compared with the free compound, BA and relevant controls.
MIA PaCa-2 cells were assessed using light microscopy to determine the effect of treatment on morphology. In this study, the characteristic spindle-shaped (elongated) morphology of MIA PaCa-2 cells was observed in the untreated sample [50]. Similarly, most BA-treated cells retained that morphology and the cells continued to grow and were still attached to neighbouring cells. A second population of smaller, rounded-up cells was also observed. It is well-documented that the rounding up and shrinkage of cells are critical characteristic features of initiation and progression into apoptosis [51,52]. This is also associated with an intact plasma membrane with reduced cytoplasm, further indicating an earlier phase of apoptosis and not necrosis [53]. Coricovac and colleagues (2021) tested the effect on the morphology of A375 human melanoma cells using BA (5-50 µM). Their results showed that apoptotic cells (smaller and rounded up) were present in a high percentage at 10 and 50 µM, suggesting that a high concentration of BA is required to visualise sufficient morphological differences to the untreated sample [54]. A similar finding was reported by Xu and colleagues (2014), where at least 30 µM of BA was required to inhibit 50% of Hela cells [55]. In contrast, PEG-BA resulted in a dose-dependent induction of cell rounding and shrinkage, which was confirmed by the cytotoxicity results even at 2 µM, suggesting that conjugation improved not only cytotoxicity but also apoptosis induction (confirmed in the downstream analysis performed in this study).
Cell cycle analysis studies, represented by the DNA content, showed an arrest of the cancer cells in the sub-G 1 phase after treatment with PEG-BA, relative to the untreated and BA-treated cells, suggesting an increase in apoptosis. Studies have shown that the presence of cells in the sub-G 1 phase indicates apoptotic cells or the induction of apoptosis [56]. This is because this phase represents hypodiploid cells with a DNA content less than that in the G 1 phase [56,57]. DNA degradation, a notable characteristic of apoptosis, can be one of the causes of the reduced DNA content, indicating that the cells found in the peak before the G 1 /G 0 are undergoing apoptosis [56,57]. BA has been shown to induce apoptosis in various tumours, including leukaemia [58], neuroblastoma [59,60], cervical cancer [61], colon cancer [62], multiple myeloma [63] and PC [8]. In this study, both PEG-BA and BA induced a reduction of cancer cells in the S-phase, suggesting inhibition of DNA synthesis. The compounds also reduced the population of cells in the G 2 /M phase, indicative of diminished cell proliferation and viability. Similar findings have been reported using PANC-1 and IGROV-1 cells treated with BA, resulting in either a sub-G 1 or a G 1 /G 0 phase arrest [64,65]. Taken together, the findings from this study suggest that the drug conjugate PEG-BA induced increased apoptosis compared with free BA, and inhibited DNA synthesis, reduced cell proliferation and viability of PC cells.
Similar to the cell cycle findings, gene expression analysis further suggested a possible apoptosis-inducing ability of PEG-BA. PEG-BA induced the overexpression of proapoptotic genes of the extrinsic and intrinsic pathways, compared with free BA. According to the Central Dogma of molecular biology, DNA serves as the genetic material that is transcribed into messenger RNA and translated into proteins which are essential for cellular functioning [66]. Therefore, a link exists between mRNA and protein levels. A study by Vogel and Marcotte (2013) showed that there is a 40% correlation between gene expression (mRNA quantification) and protein levels [67]. Hence, in this study, the real-time PCR results were compared with literature findings based on proteins that function in apoptosis to try and predict the potential apoptotic mechanism of PEG-BA in PC cells. In cancer, the expression of antiapoptotic genes is elevated, and proapoptotic genes are downregulated [20,21]. TNF and TNFSF10 were upregulated post-PEG-BA treatment in PC cells ( Figure 3A). These genes code for pro-inflammatory cytokines, which act as ligands that bind to death receptors, initiating the extrinsic pathway of apoptosis [15]. The products of TNF and TNFSF10 can be inhibited by binding to decoy receptors such as cFLIP, which are overexpressed in several cancer types and aid cell proliferation, growth and survival [26,27]. The inhibition occurs because of the absence of a death effector domain (DED), used by tumour necrosis factor receptor type 1-associated DEATH domain (TRADD) adaptor molecules and procaspase 8 to form the DISC, whose inhibition disrupts the extrinsic pathway of apoptosis [26,27]. CASPASE 2 and 8, which code for initiator caspases 2 and 8, and CASPASE 3 and 7, which code for executioner caspases 3 and 7, were also overexpressed after PEG-BA treatment ( Figure 3B,D). The initiator caspases are activated by forming complexes that facilitate their hydrolysis, which is necessary for caspase function [13]. The activated initiators then activate the executioner caspases, resulting in apoptosis activation. Both types of caspases are inhibited by the inhibitors of apoptosis (IAPs), which are generally overexpressed in cancer cells to favour survival and evade cell death [68,69]. The function of Bid protein in type 2 cells with low caspase signal is necessary for the activation of the mitochondrial pathway, either by direct inhibitory binding to the Bcl-2 protein, allowing Bax/Bak oligomerisation or directly triggering the release of Cyt C [70][71][72]. In this study, PEG-BA resulted in the overexpression of BAX and BID in the MIA PaCa-2 cells. In most cancers, including PC, both proapoptotic proteins are inhibited by high expression of Bcl-2 antiapoptotic proteins, facilitating the evasion of apoptosis [73]. A study by Naseri and colleagues (2015) showed apoptosis could be restored in cancer cells such as HepG2, T47D and HCT116 by activating Bax [74]. Conjugating PEG to BA improved the apoptosis-inducing ability of the free drug by activating proapoptotic genes.
It is well-documented that NF-κB protein subunits play an important role in inflammation, cell differentiation, proliferation, and death [75][76][77]. At rest, NF-κB is inactivated by the bound inhibitor IKB protein family preventing its translocation to the nucleus [76]. Phosphorylation inhibits IKBs and releases NF-κB, leading to dimerisation of its subunits (canonical: p50 and p65 or non-canonical: p100 and p52), which is translocated to the nucleus and activates transcription of their target genes [76,78]. In this study, NF-κB levels in the cell lysate were higher than in the supernatant, confirming its intracellular location and that disruption of the plasma membrane was necessary to release adequate amounts of the protein [76]. A slight but insignificant increase in NF-κB was observed after treatment with PEG-BA compared with untreated cancer cells. BA has been shown to activate NF-κB by inducing phosphorylation of the IKBs [6,65]. Notably, the present study showed an upregulation of TNF, an activator of NF-κB [74,75], with PEG-BA treatment. Although some studies have demonstrated the role of NF-κB in inducing cell survival pathways in several malignancies [73], several others have determined its role in apoptosis, specifically its p65 subunit [74][75][76][77]. NF-κB is necessary for BA-induced apoptotic cell death [65]. Ryan et al. (2000) demonstrated the proapoptotic role of NF-κB in Soas-2 cells. They found that p53 (a protein also increased in MIA PaCa-2 cells) induces translocation of NF-κB to the nucleus, thereby activating binding to DNA [79]. In addition, inhibiting NF-κB reduced apoptosis, suggesting that NF-κB may be crucial in the induction of apoptosis by doxorubicin in Soas-2 cells [79]. Another study showed that the activation of the NF-κB/p65 by aspirin resulted in the induction of Bax-dependent apoptosis in human colorectal carcinoma cells [80]. It is worth noting that in this study, PEG-BA treatment resulted in overexpression of BAX in the PC cells. These findings may suggest that NF-κB plays an important proapoptotic role, which may be improved by PEG-BA treatment.
PEG-BA treatment further improved antioxidant potential, demonstrated by a lower IC 50 (15.59 ± 0.64 µM), suggesting a higher radical scavenging potential than ascorbic acid and BA. On the other hand, treatment with BA resulted in an IC 50 greater than 100 µM, suggesting a lower antioxidant potential than ascorbic acid with an IC 50 of 25.58 ± 0.44 µM. A similar observation was made by Adesanwo and colleagues (2013), where they assessed the antioxidant capacity of BA isolated from Tetracera potatoria roots and reported an even higher IC 50 value of 309 µM [81]. Several studies have reported that medicinal plants possess antioxidant activities with desirable radical scavenging abilities [81,82]. This is due to the presence of phenolic groups and tocopherols that can readily donate hydrogen atoms, which is an essential aspect of the DPPH assay, as it is based on the electron transfer ability of potential antioxidants [81,83]. Therefore, with an increase in the number of hydroxyl groups (-OH) or other hydrogen donating groups (-NH, -SH, etc.) in the molecular structure of a compound, the higher the antioxidant activity [81]. Our results, therefore, demonstrate that conjugating PEG to BA improves antioxidant activities due to the inherent anti-inflammatory property of BA, but also with added advantages such as improved solubility associated with conjugation.
The DEPPD assay confirmed significantly high concentrations of hydroperoxides in the PC cells compared with the normal cells ( Figure 6), which has also been shown in other studies [32]. A moderate decrease in hydroperoxide concentrations was observed in PEG-BA-treated cancer cells compared with untreated cells. The reduction may be indicative of antioxidant activities [84]. Anticancer compounds possessing antioxidant potential are ideal due to the added advantage of combatting inflammation (ROS), potentially blocking cancer progression [31][32][33]. Our group has previously shown that PEG-BA downregulates the expression of Th1/Th2/Th17 cytokines involved in inflammatory processes, further confirming the anti-inflammatory ability of the conjugate [8,9].

Conclusions
This study has demonstrated that the potential therapeutic efficacy of betulinic acid on PC cells improved following polymer conjugation. It also showed at the morphological and molecular level that this increased efficacy is due to its ability to upregulate proapoptotic genes and, consequently, induce apoptosis while exhibiting antioxidation properties. These findings suggest that PEG-BA is further investigated as a treatment for PC using models such as 3D culture or animal models.