Polyurea Dendrimer Folate-Targeted Nanodelivery of l-Buthionine Sulfoximine as a Tool to Tackle Ovarian Cancer Chemoresistance

Ovarian cancer is a highly lethal disease, mainly due to chemoresistance. Our previous studies on metabolic remodeling in ovarian cancer have supported that the reliance on glutathione (GSH) bioavailability is a main adaptive metabolic mechanism, also accounting for chemoresistance to conventional therapy based on platinum salts. In this study, we tested the effects of the in vitro inhibition of GSH synthesis on the restoration of ovarian cancer cells sensitivity to carboplatin. GSH synthesis was inhibited by exposing cells to l-buthionine sulfoximine (l-BSO), an inhibitor of γ-glutamylcysteine ligase (GCL). Given the systemic toxicity of l-BSO, we developed a new formulation using polyurea (PURE) dendrimers nanoparticles (l-BSO@PUREG4-FA2), targeting l-BSO delivery in a folate functionalized nanoparticle.

The encapsulation of FL in PURE G4 -FA 2 (FL@PURE G4 -FA 2 ) followed the same methodology used for L-BSO encapsulation [31]. Typically, in a vial, FL (0.0131 mmol, 5.3 mg) was dissolved in 1 mL of distilled water. To this solution, the folate-target dendrimer (PURE G4 -FA 2 ) (6.46 µmol, 56.6 mg) was added. The mixture was then left overnight at RT, in the dark, and under stirring. After this period, the product was purified by dialysis (MWCO 100-500 Da) and characterized by 1 H NMR.

Confirmation of Cellular Internalization of Nanoparticles by Flow Cytometry
Cells (1 × 10 5 cells/well) were cultured overnight on 24-well plates and then incubated with several concentrations of FL@PURE G4 -FA 2 (0.001-0.120 µM) for 24 h. Only viable adherent cells were collected for the analysis. Cells were then washed with PBS (1×) and detached with trypsin-EDTA. After collection to 1.5 mL Eppendorfs, cells were harvested by centrifugation at 255× g for 3 min and washed twice with PBS (1×). Afterwards, cells were suspended in 200 µL of PBS (1×) and samples were analyzed by flow cytometry (FACScalibur-Becton Dickinson; New Jersey, NJ, USA). Sample data was analyzed using FlowJo 8.7 software (https://www.flowjo.com). The assay was performed at least in three biological replicates.

Confirmation of Cellular Internalization of Nanoparticles by Fluorescence Microscopy
The cell lines OVCAR3, ES2 and HaCaT were cultured on glass slides coated with 0.2% gelatine and then incubated with free FL or FL@PURE G4 -FA 2, for 8 and 24 h. After incubation, cells were fixed in 2% paraformaldehyde for 15 min at RT and washed with PBS (1×). The slides were mounted in VECTASHIELD media with DAPI and examined by standard fluorescence microscopy using an Axio Imager.Z1 microscope. The images were acquired with the CytoVision software. The assay was performed at least in three biological replicates.

Cell Death Analysis by Flow Cytometry
The cells (1 × 10 5 cells/well) were seeded in 24-well plates and cultured overnight in control conditions. The effect of different concentrations of free l-BSO (between 0.05 and 120 mM) and l-BSO@PURE G4 -FA 2 (between 3 and 2522 µM) in cell viability was tested for 24 h of exposure. To evaluate the sensitization effect of L-BSO to carboplatin, OVCAR3 cells were exposed to the previous culture conditions combined with carboplatin (25 µg/mL).

Statistical Analysis
Statistical analyses were performed in GraphPad Prism 7.0 software (www.graphpad.com). Data is presented as mean ± SD. Assays were performed with at least three biological replicates. For comparisons of two groups, two-tailed unpaired t-test was used. To compare more than two groups, one-way and two-way analysis of variance (ANOVA) with Dunnets multiple-comparisons test were used. Statistical significance was established at p < 0.05; * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Ovarian Cancer Cells Internalize PURE G4 -FA 2 Nanoparticles in a Dose Dependent Manner
The expression of FA-Rα was confirmed in ovarian cancer (ES2 and OVCAR3) and squamous non-cancer (HaCaT) cell lines. As seen, HaCaT cell are negative for FA-Rα, whereas ES2 and OVCAR3 cells express FA-Rα ( Figure 1A). In order to validate the specificity of the internalization of PURE G4 -FA 2 by ovarian cancer cells, we tested fluorescein loaded PURE G4 -FA 2 (FL@PURE G4 -FA 2 ) prior to test L-BSO@PURE G4 -FA 2 . By flow cytometry and fluorescence microscopy, we verified that FL is delivered in a dose dependent manner to both ES2 and OVCAR3 cell lines ( Figure 1). In HaCaT cells, the internalization of fluorescein was only verified at the highest concentration (1 µM) of FL@PURE G4 -FA 2 , after 8 and 24 h ( Figure 1C,D). This observation supports the affinity of FL@PURE G4 -FA 2 to ovarian cancer cells.
Antioxidants 2020, 9, x FOR PEER REVIEW 4 of 13 Statistical analyses were performed in GraphPad Prism 7.0 software (www.graphpad.com). Data is presented as mean ± SD. Assays were performed with at least three biological replicates. For comparisons of two groups, two-tailed unpaired t-test was used. To compare more than two groups, one-way and two-way analysis of variance (ANOVA) with Dunnets multiple-comparisons test were used. Statistical significance was established at p < 0.05; * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Ovarian Cancer Cells Internalize PUREG4-FA2 Nanoparticles in a Dose Dependent Manner
The expression of FA-Rα was confirmed in ovarian cancer (ES2 and OVCAR3) and squamous non-cancer (HaCaT) cell lines. As seen, HaCaT cell are negative for FA-Rα, whereas ES2 and OVCAR3 cells express FA-Rα ( Figure 1Α). In order to validate the specificity of the internalization of PUREG4-FA2 by ovarian cancer cells, we tested fluorescein loaded PUREG4-FA2 (FL@PUREG4-FA2) prior to test L-BSO@PUREG4-FA2. By flow cytometry and fluorescence microscopy, we verified that FL is delivered in a dose dependent manner to both ES2 and OVCAR3 cell lines ( Figure 1). In HaCaT cells, the internalization of fluorescein was only verified at the highest concentration (1 µM) of FL@PUREG4-FA2, after 8 and 24 h ( Figure 1C,D). This observation supports the affinity of FL@PUREG4-FA2 to ovarian cancer cells.

L-BSO@PUREG4-FA2 is More Effective in Inducing Cell Death in Ovarian Cancer Cells than Free L-BSO
The efficacy of inducing cell death by free and encapsulated L-BSO was tested in ovarian cancer cells, in order to verify the advantages of using a L-BSO targeted delivery. In ovarian cancer cells, the concentration needed to reach 50% of the maximum cytotoxic effect (EC50) was higher in free L-BSO than in L-BSO@PUREG4-FA2, indicating a more effective delivery of L-BSO by the L-BSO@PUREG4-FA2 nanoformulation in comparison to free L-BSO applied directly to the culture media (Figures 2 and 3

l-BSO@PURE G4 -FA 2 is More Effective in Inducing Cell Death in Ovarian Cancer Cells than Free l-BSO
The efficacy of inducing cell death by free and encapsulated l-BSO was tested in ovarian cancer cells, in order to verify the advantages of using a l-BSO targeted delivery. In ovarian cancer cells, the concentration needed to reach 50% of the maximum cytotoxic effect (EC 50 ) was higher in free L-BSO than in l-BSO@PURE G4 -FA 2 , indicating a more effective delivery of l-BSO by the l-BSO@PURE G4 -FA 2 nanoformulation in comparison to free l-BSO applied directly to the culture media (Figures 2 and 3).

L-BSO@PUREG4-FA2 Is More Cytotoxic to Ovarian Cancer Cells than to Non-Cancer Squamous Cells
L-BSO@PUREG4-FA2 was also tested in non-cancer squamous cells, as an attempt to address the effect in the peritoneal squamous cells, trying to anticipate a future therapy applied by intraabdominal infusion. Thus, the cell death levels in HaCaT cells were evaluated upon exposure to L-BSO@PUREG4-FA2, and no differences were observed between cells exposed to different concentrations of L-BSO@PUREG4-FA2 (Figure 4). Interestingly, the highest concentration tested (1000 µM) induced about 10% of cell death in squamous cells, whereas the same concentration induced more than 40% of cell death in ovarian cancer cells ( Figures 3A,B and 4). This result supports that L-BSO targeted delivery can be a good strategy to treat ovarian cancer without strongly affecting noncancer cells, at least in a re-sensitizing therapeutic protocol to overcome resistance to platinum salts. 3.3. l-BSO@PURE G4 -FA 2 Is More Cytotoxic to Ovarian Cancer Cells than to Non-Cancer Squamous Cells l-BSO@PURE G4 -FA 2 was also tested in non-cancer squamous cells, as an attempt to address the effect in the peritoneal squamous cells, trying to anticipate a future therapy applied by intra-abdominal infusion. Thus, the cell death levels in HaCaT cells were evaluated upon exposure to l-BSO@PURE G4 -FA 2 , and no differences were observed between cells exposed to different concentrations of l-BSO@PURE G4 -FA 2 ( Figure 4). Interestingly, the highest concentration tested (1000 µM) induced about 10% of cell death in squamous cells, whereas the same concentration induced more than 40% of cell death in ovarian cancer cells ( Figure 3A,B and Figure 4). This result supports that l-BSO targeted delivery can be a good strategy to treat ovarian cancer without strongly affecting non-cancer cells, at least in a re-sensitizing therapeutic protocol to overcome resistance to platinum salts.

L-BSO@PUREG4-FA2 Is Effective in Increasing the Sensitivity of Ovarian Cancer Cells to Carboplatin
To validate our re-sensitizing approach, the OVCAR3 ovarian cancer cell line was exposed to increasing concentrations of L-BSO@PUREG4-FA2 separately or combined with carboplatin. Overall, L-BSO@PUREG4-FA2 exposure improved the cytotoxic effect of carboplatin. Furthermore, L-BSO@PUREG4-FA2 by itself increased cell death, showing again the reliance of ovarian cancer cells on GSH bioavailability ( Figure 5). However, the highest concentrations of L-BSO@PUREG4-FA2 did not improve carboplatin cytotoxicity, which can be related to the threshold of cell capacity of internalizing nanoparticles.  HaCaT cells were exposed to different concentrations of l-BSO loaded into PURE G4 -FA 2 nanoparticles (L-BSO@PURE G4 -FA 2 ) and cell death percentage was determined by flow cytometry using annexin V-FITC and propidium iodide (PI) staining. Results are shown as mean ± SD.

l-BSO@PURE G4 -FA 2 Is Effective in Increasing the Sensitivity of Ovarian Cancer Cells to Carboplatin
To validate our re-sensitizing approach, the OVCAR3 ovarian cancer cell line was exposed to increasing concentrations of L-BSO@PURE G4 -FA 2 separately or combined with carboplatin. Overall, l-BSO@PURE G4 -FA 2 exposure improved the cytotoxic effect of carboplatin. Furthermore, l-BSO@PURE G4 -FA 2 by itself increased cell death, showing again the reliance of ovarian cancer cells on GSH bioavailability ( Figure 5). However, the highest concentrations of l-BSO@PURE G4 -FA 2 did not improve carboplatin cytotoxicity, which can be related to the threshold of cell capacity of internalizing nanoparticles.

L-BSO@PUREG4-FA2 Is Effective in Increasing the Sensitivity of Ovarian Cancer Cells to Carboplatin
To validate our re-sensitizing approach, the OVCAR3 ovarian cancer cell line was exposed to increasing concentrations of L-BSO@PUREG4-FA2 separately or combined with carboplatin. Overall, L-BSO@PUREG4-FA2 exposure improved the cytotoxic effect of carboplatin. Furthermore, L-BSO@PUREG4-FA2 by itself increased cell death, showing again the reliance of ovarian cancer cells on GSH bioavailability ( Figure 5). However, the highest concentrations of L-BSO@PUREG4-FA2 did not improve carboplatin cytotoxicity, which can be related to the threshold of cell capacity of internalizing nanoparticles.

Discussion
Acquired chemoresistance is a critical issue in oncology and ovarian cancer is a paradigm of this matter. Therefore, the development of strategies to overcome chemoresistance is required for a more effective treatment of ovarian cancer [8,15,32,33]. Following our insights on ovarian cancer metabolic remodeling and therapy response [11][12][13], we posited that a FA-Rα-targeted delivery of l-BSO can be a promising strategy to revoke resistance to carboplatin ( Figure 6).
We have previously shown the efficacy of PUREG4-FA2 nanoparticles as a vehicle to deliver a cytotoxic selenium-chrysin compound to ovarian cancer cells [28]. In this study, we have shown the efficacy of those nanoparticles as a vehicle to delivery also L-BSO to ovarian cancer cells. By using fluorescein loaded PUREG4-FA2, we verified that ovarian cancer cells are more competent in the internalization of these nanoparticles when compared to non-cancer cells. The observed fluorescein uptake by non-cancer squamous cells at the highest tested concentration can be explained by the use of static cultures and by the fact that after 24 h some particles can adsorb to the cells in a nonspecific way. Nevertheless, our results confirmed that the high levels of FA-Rα expression by cancer cells [29] can be explored as a way to reduce the effect of L-BSO in non-cancerous cells. Indeed, this fact allows a preservative systemic therapeutic approach, since L-BSO also induces GSH depletion [34,35] in normal cells, thus, rendering L-BSO otherwise too toxic for therapy. In the early 1990s, L-BSO was used as a drug to treat cancer [36,37], but its adverse effects were so severe that its use was promptly interrupted. However, more recently, L-BSO has regained attention, and several studies reported the use of this compound in cancer preclinical models [38][39][40][41][42]. Figure 6. Rational of L-BSO@PUREG4-FA2 sensitization to carboplatin toxicity in chemoresistant cancer cells. Cancer cells presenting a high glutathione (GSH) bioavailability are commonly resistant to platinum salts (Pt) toxicity, since GSH is a reactive oxygen species (ROS) scavenger and a xenobiotic detoxifying system. L-Buthionine sulfoximine (L-BSO) is an irreversible inhibitor of γglutamylcysteine ligase (GCL; which has catalytic and modulator subunits, GCLC and GCLM) responsible for GSH synthesis. The targeted delivery of L-BSO in folate-functionalized polyurea dendrimer generation four (PUREG4-FA2) nanoparticles, taking advantage of the increased expression of FA-Rα in cancer cells, will be efficiently internalized, inhibiting the synthesis of GSH. Therefore, carboplatin will act through its mechanisms of action, ROS generation and adducts formation, culminating in cancer cells death.
Prior to this study, we demonstrated that free L-BSO efficiently diminishes GSH bioavailability, impairing resistance to carboplatin [11]. Importantly, this effect of L-BSO was also observed in an in Figure 6. Rational of l-BSO@PURE G4 -FA 2 sensitization to carboplatin toxicity in chemoresistant cancer cells. Cancer cells presenting a high glutathione (GSH) bioavailability are commonly resistant to platinum salts (Pt) toxicity, since GSH is a reactive oxygen species (ROS) scavenger and a xenobiotic detoxifying system. l-Buthionine sulfoximine (l-BSO) is an irreversible inhibitor of α-glutamylcysteine ligase (GCL; which has catalytic and modulator subunits, GCLC and GCLM) responsible for GSH synthesis. The targeted delivery of L-BSO in folate-functionalized polyurea dendrimer generation four (PURE G4 -FA 2 ) nanoparticles, taking advantage of the increased expression of FA-Rα in cancer cells, will be efficiently internalized, inhibiting the synthesis of GSH. Therefore, carboplatin will act through its mechanisms of action, ROS generation and adducts formation, culminating in cancer cells death.
We have previously shown the efficacy of PURE G4 -FA 2 nanoparticles as a vehicle to deliver a cytotoxic selenium-chrysin compound to ovarian cancer cells [28]. In this study, we have shown the efficacy of those nanoparticles as a vehicle to delivery also l-BSO to ovarian cancer cells. By using fluorescein loaded PURE G4 -FA 2 , we verified that ovarian cancer cells are more competent in the internalization of these nanoparticles when compared to non-cancer cells. The observed fluorescein uptake by non-cancer squamous cells at the highest tested concentration can be explained by the use of static cultures and by the fact that after 24 h some particles can adsorb to the cells in a nonspecific way. Nevertheless, our results confirmed that the high levels of FA-Rα expression by cancer cells [29] can be explored as a way to reduce the effect of l-BSO in non-cancerous cells. Indeed, this fact allows a preservative systemic therapeutic approach, since l-BSO also induces GSH depletion [34,35] in normal cells, thus, rendering L-BSO otherwise too toxic for therapy. In the early 1990s, L-BSO was used as a drug to treat cancer [36,37], but its adverse effects were so severe that its use was promptly interrupted. However, more recently, L-BSO has regained attention, and several studies reported the use of this compound in cancer preclinical models [38][39][40][41][42].
Prior to this study, we demonstrated that free l-BSO efficiently diminishes GSH bioavailability, impairing resistance to carboplatin [11]. Importantly, this effect of L-BSO was also observed in an in vivo model of ovarian cancer, reducing significantly subcutaneous tumor size and GSH levels, as well as peritoneal dissemination [11]. In the present study, we verified that a l-BSO@PURE G4 -FA 2 nanoformulation (Figure 7) is more effective in inducing ovarian cancer cells death than free l-BSO; and that ovarian cancer cells are more sensitive to l-BSO@PURE G4 -FA 2 than non-cancer squamous cells (HaCaT), reinforcing a putative therapy mediated by abdominal infusion.
Antioxidants 2020, 9, x FOR PEER REVIEW 10 of 13 vivo model of ovarian cancer, reducing significantly subcutaneous tumor size and GSH levels, as well as peritoneal dissemination [11]. In the present study, we verified that a L-BSO@PUREG4-FA2 nanoformulation (Figure 7) is more effective in inducing ovarian cancer cells death than free L-BSO; and that ovarian cancer cells are more sensitive to L-BSO@PUREG4-FA2 than non-cancer squamous cells (HaCaT), reinforcing a putative therapy mediated by abdominal infusion. Our previous studies suggested a stronger dependence of ES2 cells on GSH turnover compared with OVCAR3 cells [12,13], which was also evidenced in this study, as higher EC50 of free L-BSO and L-BSO@PUREG4-FA2 were determined for ES2 compared with OVCAR3 cells. Furthermore, concerning resistance to carboplatin, we have reported that upon carboplatin exposure ES2 produce higher levels of GSH [11] together with an accelerated GSH turnover, compared with OVCAR3 [28]. Therefore, our greatest achievements in this study were the effective use of L-BSO@PUREG4-FA2 nanoparticles to the specific targeting of malignant cells, decreasing the harmful effects of L-BSO in non-malignant cells, and the similar effective targeting of ovarian cancer cells with different levels of chemoresistance. Together, our study supports the use of L-BSO@PUREG4-FA2 nanoparticles as a powerful strategy for ovarian cancer treatment.

Conclusions
More validation studies, namely in vivo assays, aiming to evaluate the systemic cytotoxic effect of L-BSO@PUREG4-FA2, are needed. Nevertheless, our study points out this new nanoformulation as a way of avoiding L-BSO systemic toxicity, and as a tool to abolish cancer cells resistance to carboplatin or putatively to other alkylating/oxidative drugs. In the future, this approach may be applied to other chemoresistant cancers.  Our previous studies suggested a stronger dependence of ES2 cells on GSH turnover compared with OVCAR3 cells [12,13], which was also evidenced in this study, as higher EC 50 of free l-BSO and l-BSO@PURE G4 -FA 2 were determined for ES2 compared with OVCAR3 cells. Furthermore, concerning resistance to carboplatin, we have reported that upon carboplatin exposure ES2 produce higher levels of GSH [11] together with an accelerated GSH turnover, compared with OVCAR3 [28]. Therefore, our greatest achievements in this study were the effective use of l-BSO@PURE G4 -FA 2 nanoparticles to the specific targeting of malignant cells, decreasing the harmful effects of l-BSO in non-malignant cells, and the similar effective targeting of ovarian cancer cells with different levels of chemoresistance. Together, our study supports the use of l-BSO@PURE G4 -FA 2 nanoparticles as a powerful strategy for ovarian cancer treatment.

Conclusions
More validation studies, namely in vivo assays, aiming to evaluate the systemic cytotoxic effect of l-BSO@PURE G4 -FA 2 , are needed. Nevertheless, our study points out this new nanoformulation as a way of avoiding l-BSO systemic toxicity, and as a tool to abolish cancer cells resistance to carboplatin or putatively to other alkylating/oxidative drugs. In the future, this approach may be applied to other chemoresistant cancers.