*2.1. BUD:HP*β*CD Complex and HP*β*CD Attenuate H2O<sup>2</sup>* + *LPS-Induced Cytotoxicity*

suffering from smoking-induced COPD.

Since alveolar cell death is one feature observed in the lung of patients suffering from smoking-induced COPD [31], the potential effect of BUD:HPβCD on A549 human alveolar epithelial cells submitted to oxidant and inflammatory stressors was investigated. Cells were incubated with H2O<sup>2</sup> + LPS for 2 h. Cytotoxicity, as reflected by lactate dehydrogenase (LDH) release, was observed with a 1.7-fold increase as compared to untreated cells (Figure 1A). The increase in cytotoxicity between 2 and 6 h is low (1.8-fold at 6 h) suggesting the cytotoxicity almost reached its maximum at 2 h. H2O<sup>2</sup> + LPS-induced cytotoxicity seemed to result from the addition of H2O<sup>2</sup> and LPS. *2.1. BUD:HPβCD Complex and HPβCD Attenuate H2O2 + LPS-Induced Cytotoxicity*  Since alveolar cell death is one feature observed in the lung of patients suffering from smokinginduced COPD [31], the potential effect of BUD:HPβCD on A549 human alveolar epithelial cells submitted to oxidant and inflammatory stressors was investigated. Cells were incubated with H2O2 + LPS for 2 h. Cytotoxicity, as reflected by lactate dehydrogenase (LDH) release, was observed with a 1.7-fold increase as compared to untreated cells (Figure 1A). The increase in cytotoxicity between 2 and 6 h is low (1.8-fold at 6 h) suggesting the cytotoxicity almost reached its maximum at 2 h. H2O2 + LPS-induced cytotoxicity seemed to result from the addition of H2O2 and LPS.

The effect of the BUD:HPβCD complex on cytotoxicity induced by H2O<sup>2</sup> + LPS was followed. Incubation of A549 cells with the BUD:HPβCD complex together with H2O<sup>2</sup> + LPS induced a decrease in cytotoxicity (Figure 1D). This protective effect appeared to not evolve further after 2 h of incubation. A similar effect was recorded with HPβCD (Figure 1C), whereas BUD showed no effect whatever the dosage (Figure 1B). These results suggest that the BUD:HPβCD complex and HPβCD would have anticytotoxic potential against H2O<sup>2</sup> + LPS-induced cytotoxicity in A549 cells. The effect of the BUD:HPβCD complex on cytotoxicity induced by H2O2 + LPS was followed. Incubation of A549 cells with the BUD:HPβCD complex together with H2O2 + LPS induced a decrease in cytotoxicity (Figure 1D). This protective effect appeared to not evolve further after 2h of incubation. A similar effect was recorded with HPβCD (Figure 1C), whereas BUD showed no effect whatever the dosage (Figure 1B). These results suggest that the BUD:HPβCD complex and HPβCD would have anticytotoxic potential against H2O2 + LPS-induced cytotoxicity in A549 cells.

**Figure 1.** Lactate dehydrogenase (LDH) release after A549 cells incubation with H2O2, Lipopolysaccharides (LPSs), and H2O2 + LPS for up to 6 h (**A**) and effect of budesonide (BUD) (**B**), **Figure 1.** Lactate dehydrogenase (LDH) release after A549 cells incubation with H2O<sup>2</sup> , Lipopolysaccharides (LPSs), and H2O<sup>2</sup> + LPS for up to 6 h (**A**) and effect of budesonide (BUD) (**B**), HPβCD (**C**), and BUD:HPβCD complex (**D**) on H2O<sup>2</sup> + LPS-induced cytotoxicity. Each point represents the mean ± SEM of at least 4 independent means of triplicated measures; where not visible, error bars are included in the symbol. The difference was considered significant for a *p*-value < 0.05. (aaa) indicates *p* < 0.001 versus untreated group; (\*\*) and (\*\*\*) corresponds to *p* < 0.01 and 0.001 versus H2O<sup>2</sup> +LPS-treated group, respectively.

To determine if apoptosis is involved in the cell death process for which BUD:HPβCD could protect, apoptosis was monitored by counting condensed/fragmented nuclei using HOECHST dye on H2O<sup>2</sup> + LPS-treated cells. We also determined if the BUD:HPβCD complex as well as BUD, and HPβCD could attenuate apoptosis. To determine if apoptosis is involved in the cell death process for which BUD:HPβCD could protect, apoptosis was monitored by counting condensed/fragmented nuclei using HOECHST dye on H2O2 + LPS-treated cells. We also determined if the BUD:HPβCD complex as well as BUD, and HPβCD could attenuate apoptosis.

*Molecules* **2020**, *25*, x 4 of 21

HPβCD (**C**), and BUD:HPβCD complex (**D**) on H2O2 + LPS-induced cytotoxicity. Each point represents the mean ± SEM of at least 4 independent means of triplicated measures; where not visible, error bars are included in the symbol. The difference was considered significant for a *p*-value < 0.05. (aaa) indicates *p* < 0.001 versus untreated group; (\*\*) and (\*\*\*) corresponds to *p* < 0.01 and 0.001 versus

A549 cells were incubated for 2 h with H2O<sup>2</sup> + LPS with/without BUD:HPβCD in comparison with BUD or HPβCD. H2O<sup>2</sup> + LPS induced significant apoptosis, which appeared to result from the addition of the individual effects of H2O<sup>2</sup> and LPS (Figure 2A). Concomitant incubation with each of the selected compounds induced a concentration-dependent decrease in H2O<sup>2</sup> + LPS-induced apoptosis. These results suggest that the BUD:HPβCD complex (Figure 2D) and HPβCD (Figure 2C), at the highest selected dosages could protect cells against H2O<sup>2</sup> + LPS-induced apoptosis in A549 cells. A non-significant decrease was observed with BUD (Figure 2B). A549 cells were incubated for 2 h with H2O2 + LPS with/without BUD:HPβCD in comparison with BUD or HPβCD. H2O2 + LPS induced significant apoptosis, which appeared to result from the addition of the individual effects of H2O2 and LPS (Figure 2A). Concomitant incubation with each of the selected compounds induced a concentration-dependent decrease in H2O2 + LPS-induced apoptosis. These results suggest that the BUD:HPβCD complex (Figure 2D) and HPβCD (Figure 2C), at the highest selected dosages could protect cells against H2O2 + LPS-induced apoptosis in A549 cells. A non-significant decrease was observed with BUD (Figure 2B).

**Figure 2.** A549 cells apoptosis after treatment with H2O2, LPS, and H2O2 + LPS for 2 h (**A**) and effect of BUD (**B**), HPβCD (**C**), and BUD:HPβCD complex (**D**) on H2O2 + LPS-induced apoptosis. Apoptosis was quantified by counting condensed/fragmented nuclei after HOECHST staining. Each bar represents the mean of 3 ± SEM or 2 independent measures. A one-way ANOVA with Dunett posttest was used to compare the mean of a test group with the mean of the untreated group or H2O2 +LPS-treated group. The difference was considered significant for a *p*-value < 0.05. (aa) indicates *p* < 0.01 versus untreated group, (\*) indicates *p* < 0.05 versus H2O2 +LPS-treated group. **Figure 2.** A549 cells apoptosis after treatment with H2O<sup>2</sup> , LPS, and H2O<sup>2</sup> + LPS for 2 h (**A**) and effect of BUD (**B**), HPβCD (**C**), and BUD:HPβCD complex (**D**) on H2O<sup>2</sup> + LPS-induced apoptosis. Apoptosis was quantified by counting condensed/fragmented nuclei after HOECHST staining. Each bar represents the mean of 3 ± SEM or 2 independent measures. A one-way ANOVA with Dunett post-test was used to compare the mean of a test group with the mean of the untreated group or H2O<sup>2</sup> +LPS-treated group. The difference was considered significant for a *p*-value < 0.05. (aa) indicates *p* < 0.01 versus untreated group, (\*) indicates *p* < 0.05 versus H2O<sup>2</sup> +LPS-treated group.

#### *2.2. BUD:HPβCD Complex and HPβCD Protect against H2O2 + LPS-Induced Oxidative Stress in A549 2.2. BUD:HP*β*CD Complex and HP*β*CD Protect against H2O<sup>2</sup>* + *LPS-Induced Oxidative Stress in A549 Cells: Dose and Time-Dependent E*ff*ects*

*Cells: Dose and Time-Dependent Effects*  Because oxidative stress is critical for numerous pathologies including smoking-induced COPD [32], and with the aim to understand the mechanism of action behind the effects observed with BUD:HPβCD, the potential antioxidant effect of the BUD:HPβCD complex in A549 human alveolar epithelial cells was monitored. A549 cells were incubated for 2 h with a H2O2 + LPS mix to model a Because oxidative stress is critical for numerous pathologies including smoking-induced COPD [32], and with the aim to understand the mechanism of action behind the effects observed with BUD:HPβCD, the potential antioxidant effect of the BUD:HPβCD complex in A549 human alveolar epithelial cells was monitored. A549 cells were incubated for 2 h with a H2O<sup>2</sup> + LPS mix to model a concomitant oxidative and inflammatory environment. ROS generation induced by the BUD:HPβCD complex was monitored in comparison with the effect induced by budesonide or HPβCD.

In comparison with control cells, we observed a 1.6-fold significant increase in intracellular ROS production when cells were incubated with H2O<sup>2</sup> + LPS (Figure 3A) This effect was similar to the effect induced by treatment with H2O<sup>2</sup> alone (1.5-fold increase), unlike the treatment with LPS alone, would be mainly driven by H2O2 in A549 cells.

which did not show any significant effects, suggesting that H2O<sup>2</sup> + LPS-induced oxidative stress would be mainly driven by H2O<sup>2</sup> in A549 cells. H2O2 + LPS. A similar effect was observed with increasing concentrations of HPβCD (25–2500 µM; Figure 3C). In contrast no effect of BUD (1–100 µM; Figure 3B) was observed.

This suggests a protective effect of the BUD:HPβCD complex against the oxidative stress induced by

*Molecules* **2020**, *25*, x 5 of 21

concomitant oxidative and inflammatory environment. ROS generation induced by the BUD:HPβCD

In comparison with control cells, we observed a 1.6-fold significant increase in intracellular ROS production when cells were incubated with H2O2 + LPS (Figure 3A) This effect was similar to the effect induced by treatment with H2O2 alone (1.5-fold increase), unlike the treatment with LPS alone, which did not show any significant effects, suggesting that H2O2 + LPS-induced oxidative stress

A concomitant incubation of A549 cells with H2O2 + LPS and increasing concentrations of the BUD:HPβCD complex (1:25, 10:250, 100:2500 µM; Figure 3D) was associated with a decrease in ROS

complex was monitored in comparison with the effect induced by budesonide or HPβCD.

**Figure 3.** ROS generation in A549 cells after treatment with H2O2, LPS, or H2O2 + LPS for 2 h (**A**) and effect of BUD (**B**), HPβCD (**C**), and BUD:HPβCD complex (**D**) on H2O2 + LPS-induced ROS generation. ROS generation was evaluated by measuring the fluorescence of dichlorofluorescein (DCF). Each bar represents the mean ± SEM of 4 independent means of triplicated measures. A one-way ANOVA with Dunett post-test was used to compare the mean of a test group with the mean of untreated group or H2O2 + LPS-treated group. The difference was considered significant for a *p*-value < 0.05. (aa), and (aaa), indicate *p* < 0.01, and 0.001 versus untreated group, respectively); (\*\*), and (\*\*\*), correspond to **Figure 3.** ROS generation in A549 cells after treatment with H2O<sup>2</sup> , LPS, or H2O<sup>2</sup> + LPS for 2 h (**A**) and effect of BUD (**B**), HPβCD (**C**), and BUD:HPβCD complex (**D**) on H2O<sup>2</sup> + LPS-induced ROS generation. ROS generation was evaluated by measuring the fluorescence of dichlorofluorescein (DCF). Each bar represents the mean ± SEM of 4 independent means of triplicated measures. A one-way ANOVA with Dunett post-test was used to compare the mean of a test group with the mean of untreated group or H2O<sup>2</sup> + LPS-treated group. The difference was considered significant for a *p*-value < 0.05. (aa), and (aaa), indicate *p* < 0.01, and 0.001 versus untreated group, respectively); (\*\*), and (\*\*\*), correspond to *p* < 0.01 and 0.001 versus H2O<sup>2</sup> + LPS-treated group, respectively.

*p* < 0.01 and 0.001 versus H2O2 + LPS-treated group, respectively. Regarding the effect of time (Figure 4), ROS production in the presence of H2O2 + LPS was already marked after 30 min compared to untreated cells. This effect was maintained throughout the entire time period investigated (6h) and seemed to evolve in parallel with untreated cells after 2 h. When the BUD:HPβCD complex was added together with H2O2 + LPS, a lowering effect on ROS production was observed throughout the entire time period investigated (Figure 4B). The extent of A concomitant incubation of A549 cells with H2O<sup>2</sup> + LPS and increasing concentrations of the BUD:HPβCD complex (1:25, 10:250, 100:2500 µM; Figure 3D) was associated with a decrease in ROS production as compared with the experimental conditions in which the complex was not present. This suggests a protective effect of the BUD:HPβCD complex against the oxidative stress induced by H2O<sup>2</sup> + LPS. A similar effect was observed with increasing concentrations of HPβCD (25–2500 µM; Figure 3C). In contrast no effect of BUD (1–100 µM; Figure 3B) was observed.

the effect depended upon the dose and was largely similar to the effect observed in the presence of HPβCD (Figure 4C). No significant change was observed after 2 h of incubation with H2O2 + LPS. Again, during the entire period investigated, no significant effect was observed in the presence of BUD (Figure 4A). Altogether, these results suggest that the BUD:HPβCD complex and HPβCD have a similar antioxidant potential against H2O2 + LPS in A549 cells. Regarding the effect of time (Figure 4), ROS production in the presence of H2O<sup>2</sup> + LPS was already marked after 30 min compared to untreated cells. This effect was maintained throughout the entire time period investigated (6 h) and seemed to evolve in parallel with untreated cells after 2 h. When the BUD:HPβCD complex was added together with H2O<sup>2</sup> + LPS, a lowering effect on ROS production was observed throughout the entire time period investigated (Figure 4B). The extent of the effect depended upon the dose and was largely similar to the effect observed in the presence of HPβCD (Figure 4C). No significant change was observed after 2 h of incubation with H2O<sup>2</sup> + LPS. Again, during the entire period investigated, no significant effect was observed in the presence of BUD (Figure 4A). Altogether, these results suggest that the BUD:HPβCD complex and HPβCD have a similar antioxidant potential against H2O<sup>2</sup> + LPS in A549 cells.

similar way in A549 cells.

**Figure 4.** Effect of BUD (**A**), BUD:HPβCD complex (**B**) and HPβCD (**C**) on H2O2 + LPS-induced ROS generation for 0 to 6h of incubation. ROS generation was evaluated by measuring the fluorescence of dichlorofluorescein (DCF). Each bar represents the mean ± SEM of 3 independent means of triplicated measures. These results and the results illustrated in Figure 3 are independent. When deviations are not visible they are too small to be seen. The difference was considered significant for a *p*-value < 0.05. (aaa) corresponds to *p* < 0.001 versus untreated group; (\*\*\*) indicates *p* < 0.001 versus H2O2 + LPS-**Figure 4.** Effect of BUD (**A**), BUD:HPβCD complex (**B**) and HPβCD (**C**) on H2O<sup>2</sup> + LPS-induced ROS generation for 0 to 6h of incubation. ROS generation was evaluated by measuring the fluorescence of dichlorofluorescein (DCF). Each bar represents the mean ± SEM of 3 independent means of triplicated measures. These results and the results illustrated in Figure 3 are independent. When deviations are not visible they are too small to be seen. The difference was considered significant for a *p*-value < 0.05. (aaa) corresponds to *p* < 0.001 versus untreated group; (\*\*\*) indicates *p* < 0.001 versus H2O<sup>2</sup> + LPS-treated group.

#### treated group. *2.3. BUD:HPβCD Complex and HPβCD Attenuate H2O2 + LPS-Induced Phosphoinositide-3-Kinase/Akt 2.3. BUD:HP*β*CD Complex and HP*β*CD Attenuate H2O<sup>2</sup>* + *LPS-Induced Phosphoinositide-3-Kinase*/*Akt Signaling in A549 Cells*

*Signaling in A549 Cells*  Oxidative stress-induced glucocorticoid insensitivity involves an increase in PI3K/Akt signaling [8,33,34] as reflected by Akt phosphorylation. To validate in in vitro model the relationship between oxidative stress and increase in PI3K/Akt signaling, the phosphorylation of Akt, in the absence or in the presence of antioxidants (N-acetyl-L-cysteine (NAC), vitamin C (Vit C)) and of a PI3K inhibitor (LY294002) (Figure 5A,C) was measured. An incubation of A549 cells with H2O2 + LPS for 2 h increased Akt phosphorylation. Concomitant incubation with NAC or vitamin C or LY294002 was associated with a lower phosphorylation of Akt (of approximately 36% (NAC) and 65% (Vit C)) or a complete suppression of phosphorylation (LY294002) (Figure 5A,C). Regarding the effect of the BUD:HPβCD complex or of HPβCD at a concentration at which a significant antioxidant effect was observed, we demonstrated a decrease in Akt phosphorylation (Figure 5B,D). The decrease was approximately 32% both for the BUD:HPβCD complex and for HPβCD (Figure 5B,D). Thus, the BUD:HPβCD complex and HPβCD inhibited H2O2 + LPS-induced PI3K/Akt signaling increases in a Oxidative stress-induced glucocorticoid insensitivity involves an increase in PI3K/Akt signaling [8,33,34] as reflected by Akt phosphorylation. To validate in in vitro model the relationship between oxidative stress and increase in PI3K/Akt signaling, the phosphorylation of Akt, in the absence or in the presence of antioxidants (N-acetyl-L-cysteine (NAC), vitamin C (Vit C)) and of a PI3K inhibitor (LY294002) (Figure 5A,C) was measured. An incubation of A549 cells with H2O2 + LPS for 2 h increased Akt phosphorylation. Concomitant incubation with NAC or vitamin C or LY294002 was associated with a lower phosphorylation of Akt (of approximately 36% (NAC) and 65% (Vit C)) or a complete suppression of phosphorylation (LY294002) (Figure 5A,C). Regarding the effect of the BUD:HPβCD complex or of HPβCD at a concentration at which a significant antioxidant effect was observed, we demonstrated a decrease in Akt phosphorylation (Figure 5B,D). The decrease was approximately 32% both for the BUD:HPβCD complex and for HPβCD (Figure 5B,D). Thus, the BUD:HPβCD complex and HPβCD inhibited H2O2 + LPS-induced PI3K/Akt signaling increases in a similar way in A549 cells.

*Molecules* **2020**, *25*, x 7 of 21

**Figure 5.** Akt phosphorylation induced by H2O2 + LPS in A549 cells. Effect of *N*-acetyl-L-cysteine (NAC), vitamin C (VitC) and LY294002 (**A** and **C**) and BUD:HPβCD complex and HPβCD (**B** and **D**) after 2 h of incubation. Data are expressed in absolute values (A/B; with representative blots) or in relative values (in comparison with the pAkt/Akt ration of cells incubated with H2O2 + LPS; C/D). Akt phosphorylation was quantified after a Western blot by measuring the proportion of phosphorylated-Akt (p-Akt) blot luminescence intensity/total Akt (Akt) blot luminescence intensity. Each bar represents the mean of 3 independent measures. A one-way ANOVA with Dunett post-test was used to compare the mean of each test group with the mean of untreated group or H2O2 + LPS-treated group. The difference was considered significant for a *p*-value < 0.05. (aaa) correspond to *p* < 0.001 versus untreated group, (\*) and (\*\*\*) indicate *p* < 0.05, and 0.001 versus H2O2 + LPS-treated group, **Figure 5.** Akt phosphorylation induced by H2O<sup>2</sup> + LPS in A549 cells. Effect of *N*-acetyl-L-cysteine (NAC), vitamin C (VitC) and LY294002 (**A**,**C**) and BUD:HPβCD complex and HPβCD (**B**,**D**) after 2 h of incubation. Data are expressed in absolute values (A/B; with representative blots) or in relative values (in comparison with the pAkt/Akt ration of cells incubated with H2O<sup>2</sup> + LPS; C/D). Akt phosphorylation was quantified after a Western blot by measuring the proportion of phosphorylated-Akt (p-Akt) blot luminescence intensity/total Akt (Akt) blot luminescence intensity. Each bar represents the mean of 3 independent measures. A one-way ANOVA with Dunett post-test was used to compare the mean of each test group with the mean of untreated group or H2O<sup>2</sup> + LPS-treated group. The difference was considered significant for a *p*-value < 0.05. (aaa) correspond to *p* < 0.001 versus untreated group, (\*) and (\*\*\*) indicate *p* < 0.05, and 0.001 versus H2O<sup>2</sup> + LPS-treated group, respectively.

#### respectively. *2.4. Cholesterol Might Limit the Effects of BUD:HPβCD Complex and HPβCD in ROS Generation and 2.4. Cholesterol Might Limit the E*ff*ects of BUD:HP*β*CD Complex and HP*β*CD in ROS Generation and PI3K*/*Akt Signaling Induced by H2O<sup>2</sup>* + *LPS*

*PI3K/Akt Signaling Induced by H2O2 + LPS*  To give insight on the molecular mechanisms involved in the protective effect of BUD:HPBCD and HPBCD, the potential role of cholesterol on ROS generation and PI3/Akt phosphorylation induced by H2O2 + LPS as well as the protective effects of the BUD:HPβCD complex and HPβCD were investigated. The rationale was derived from the ability of cyclodextrins to interact with cholesterol [35], the effects of BUD:HPβCD and HPβCD on the biophysical membrane properties of cholesterol-enriched domains [27], the importance of lipid-ordered domains enriched in cholesterol To give insight on the molecular mechanisms involved in the protective effect of BUD:HPBCD and HPBCD, the potential role of cholesterol on ROS generation and PI3/Akt phosphorylation induced by H2O<sup>2</sup> + LPS as well as the protective effects of the BUD:HPβCD complex and HPβCD were investigated. The rationale was derived from the ability of cyclodextrins to interact with cholesterol [35], the effects of BUD:HPβCD and HPβCD on the biophysical membrane properties of cholesterol-enriched domains [27], the importance of lipid-ordered domains enriched in cholesterol in membrane called rafts for ROS generation [36,37] and PI3K/Akt signaling [23,26].

in membrane called rafts for ROS generation [36,37] and PI3K/Akt signaling [23,26].

ROS Generation Induced by H2O2 + LPS

2.4.1. Cholesterol Content Might Influence the Effects of the BUD:HPβCD Complex and HPβCD in ROS Generation Induced by H2O<sup>2</sup> + LPS Compared to non-depleted cells, the ability of the BUD:HPβCD complex (Figure 6C) and HPβCD (Figure 6D) to protect against H2O2 + LPS-induced ROS production was preserved. Moreover, when the difference between H2O2 + LPS-induced ROS production in cholesterol-depleted

cholesterol content plays a role in H2O2 + LPS-related oxidative signaling in A549 cells.

*Molecules* **2020**, *25*, x 8 of 21

2.4.1. Cholesterol Content Might Influence the Effects of the BUD:HPβCD Complex and HPβCD in

In conditions where cholesterol was partly depleted (see Figure S1) we observed (Figure 6A) an increase in basal intracellular ROS levels, although non-significant. A greater and significant increase

In conditions where cholesterol was partly depleted (see Figure S1) we observed (Figure 6A) an increase in basal intracellular ROS levels, although non-significant. A greater and significant increase in H2O2- and H2O<sup>2</sup> + LPS-induced intracellular oxidant generation was also observed, suggesting that cholesterol content plays a role in H2O<sup>2</sup> + LPS-related oxidative signaling in A549 cells. and non-depleted cells was considered, a concentration-dependent increase in this protective effect was observed (Figure 6E,F), although this was non-significant. Again, no matter the cholesterol status of the cells, budesonide did not show any protective effects (Figure 6B). Thus, cholesterol content might influence the antioxidant effect of the BUD:HPβCD complex and HPβCD.

**Figure 6.** ROS generation in cholesterol-non-depleted or cholesterol-depleted A549 cells after treatment with H2O2, LPS, or H2O2 + LPS for 2 h (**A**) and effect of BUD (**B**), BUD:HPβCD complex (**C**) and HPβCD (**D**) on H2O2 + LPS-induced ROS generation. Panels E and F show the difference (∆) between the effect observed in cholesterol-depleted and -non-depleted cells of BUD:HPβCD complex (**E**) and HPβCD (**F**) on H2O2 + LPS-induced oxidant generation after treatment for 2 h. Each bar represents the mean ± SEM of 4 independent means of triplicated measures. H2O2 + LPS-treated bar is the same for each panel. A one-way ANOVA with Dunett post-test was used to compare the mean of each test group in panels (**E**) and (**F**) with the mean of the H2O2 + LPS-treated group. A two-way ANOVA with Tukey multiple comparison post-test was used to compare the mean of non-depleted group with the mean of cholesterol-depleted group in the same concentration (panels **A**–**D**). The **Figure 6.** ROS generation in cholesterol-non-depleted or cholesterol-depleted A549 cells after treatment with H2O<sup>2</sup> , LPS, or H2O<sup>2</sup> + LPS for 2 h (**A**) and effect of BUD (**B**), BUD:HPβCD complex (**C**) and HPβCD (**D**) on H2O<sup>2</sup> + LPS-induced ROS generation. Panels E and F show the difference (∆) between the effect observed in cholesterol-depleted and -non-depleted cells of BUD:HPβCD complex (**E**) and HPβCD (**F**) on H2O<sup>2</sup> + LPS-induced oxidant generation after treatment for 2 h. Each bar represents the mean ± SEM of 4 independent means of triplicated measures. H2O<sup>2</sup> + LPS-treated bar is the same for each panel. A one-way ANOVA with Dunett post-test was used to compare the mean of each test group in panels (**E**,**F**) with the mean of the H2O<sup>2</sup> + LPS-treated group. A two-way ANOVA with Tukey multiple comparison post-test was used to compare the mean of non-depleted group with the mean of cholesterol-depleted group in the same concentration (panels **A**–**D**). The difference was considered significant for a *p*-value < 0.05. (\*) and (\*\*) indicate, respectively, *p* < 0.05 and 0.01 between non-depleted and cholesterol-depleted group (panels **A**–**D**).

Compared to non-depleted cells, the ability of the BUD:HPβCD complex (Figure 6C) and HPβCD (Figure 6D) to protect against H2O<sup>2</sup> + LPS-induced ROS production was preserved. Moreover, when the difference between H2O<sup>2</sup> + LPS-induced ROS production in cholesterol-depleted and non-depleted cells was considered, a concentration-dependent increase in this protective effect was observed (Figure 6E,F), although this was non-significant. Again, no matter the cholesterol status of the cells, budesonide did not show any protective effects (Figure 6B). Thus, cholesterol content might influence the antioxidant effect of the BUD:HPβCD complex and HPβCD. difference was considered significant for a *p*-value < 0.05. (\*) and (\*\*) indicate, respectively, *p* < 0.05 and 0.01 between non-depleted and cholesterol-depleted group (panels **A**–**D**).

*Molecules* **2020**, *25*, x 9 of 21

In contrast with cholesterol, sphingomyelin, another major component from raft and also interacting with HPβCD did not play a critical role on neither the oxidant generation induced by H2O<sup>2</sup> + LPS nor on the ability of the BUD:HPβCD complex and HPβCD to protect against H2O<sup>2</sup> + LPS-induced oxidant generation (Figure S2). In contrast with cholesterol, sphingomyelin, another major component from raft and also interacting with HPβCD did not play a critical role on neither the oxidant generation induced by H2O2 + LPS nor on the ability of the BUD:HPβCD complex and HPβCD to protect against H2O2 + LPSinduced oxidant generation (Figure S2).

2.4.2. Cholesterol Limits the Effects of the BUD:HPβCD Complex and HPβCD in PI3K/Akt Signaling Induced by H2O<sup>2</sup> + LPS 2.4.2. Cholesterol limits the effects of the BUD:HPβCD complex and HPβCD in PI3K/Akt signaling induced by H2O2 + LPS

Since cholesterol-enriched plasma membrane domains may also play a critical role in the activation of PI3K/Akt signaling [23,25,26], which could be modulated by oxidative stress, the effect of cholesterol depletion on the protective effects of the BUD:HPβCD complex and HPβCD on the phosphorylation of Akt was studied. A decrease in Akt phosphorylation was preserved (Figure 7), without difference in cholesterol-depleted or not depleted cells. Since cholesterol-enriched plasma membrane domains may also play a critical role in the activation of PI3K/Akt signaling [23,25,26], which could be modulated by oxidative stress, the effect of cholesterol depletion on the protective effects of the BUD:HPβCD complex and HPβCD on the phosphorylation of Akt was studied. A decrease in Akt phosphorylation was preserved (Figure 7), without difference in cholesterol-depleted or not depleted cells.

**Figure 7.** Effect of the BUD:HPβCD complex versus HPβCD on H2O2 + LPS-induced Akt phosphorylation (p-Akt) in cholesterol-depleted and non-depleted A549 cells after 2 h of incubation with representative blot (cholesterol-depleted cells); each bar represents the mean of 3 independent **Figure 7.** Effect of the BUD:HPβCD complex versus HPβCD on H2O<sup>2</sup> + LPS-induced Akt phosphorylation (p-Akt) in cholesterol-depleted and non-depleted A549 cells after 2 h of incubation with representative blot (cholesterol-depleted cells); each bar represents the mean of 3 independent measures.

#### measures. *2.5. BUD:HPβCD Complex and HPβCD Protect against H2O2 + LPS-Induced Increase in HDAC2 2.5. BUD:HP*β*CD Complex and HP*β*CD Protect against H2O<sup>2</sup>* + *LPS-Induced Increase in HDAC2 Phosphorylation in A549 Cells*

*Phosphorylation in A549 cells*  The increase in PI3K/Akt signaling induced by oxidative stress results in the phosphorylation of HDAC2, a critical step in oxidative stress-related glucocorticoid insensitivity [38,39]. The relationship between oxidative stress and HDAC2 phosphorylation as well as the relationship between the increase in PI3K/Akt signaling and HDAC2 phosphorylation was investigated by using NAC and LY294002, respectively (Figure 8A,C). The treatment of A549 cells with H2O2 + LPS for 2 h was associated with an increase in phosphorylated HDAC2. Concomitant incubation with NAC or LY294002 was associated with a lower phosphorylation of HDAC2 of approximately 62% (NAC) and The increase in PI3K/Akt signaling induced by oxidative stress results in the phosphorylation of HDAC2, a critical step in oxidative stress-related glucocorticoid insensitivity [38,39]. The relationship between oxidative stress and HDAC2 phosphorylation as well as the relationship between the increase in PI3K/Akt signaling and HDAC2 phosphorylation was investigated by using NAC and LY294002, respectively (Figure 8A,C). The treatment of A549 cells with H2O<sup>2</sup> + LPS for 2 h was associated with an increase in phosphorylated HDAC2. Concomitant incubation with NAC or LY294002 was associated with a lower phosphorylation of HDAC2 of approximately 62% (NAC) and 40% (LY294002) (Figure 8A,C). This confirms that H2O<sup>2</sup> + LPS increases HDAC2 phosphorylation through a mechanism involving oxidative stress and PI3K/Akt signaling in A549 cells.

40% (LY294002) (Figure 8A,C). This confirms that H2O2 + LPS increases HDAC2 phosphorylation through a mechanism involving oxidative stress and PI3K/Akt signaling in A549 cells. A concomitant incubation with the BUD:HPβCD complex or HPβCD with H2O2 + LPS was A concomitant incubation with the BUD:HPβCD complex or HPβCD with H2O<sup>2</sup> + LPS was associated with a lower phosphorylation of HDAC2 of approximately 53% (BUD:HPβCD) and 74%

associated with a lower phosphorylation of HDAC2 of approximately 53% (BUD:HPβCD) and 74% (HPβCD) (Figure 8B,D). Thus, the BUD:HPβCD complex and HPβCD inhibited the decrease in

*Molecules* **2020**, *25*, x 10 of 21

**Figure 8.** HDAC2 phosphorylation induced by H2O2 + LPS in A549 cells. Effect of NAC and LY294002 (**A**) and the BUD:HPβCD complex versus HPβCD (**B**) after 2 h of incubation. Data are expressed in absolute values (**A**/**B**; with representative blots) or in relative values (in comparison with the pHDAC2/HDAC2 ration of cells incubated with H2O2 + LPS; **C**/**D**). HDAC2 phosphorylation was quantified after a Western blot by measuring the proportion of phosphorylated-HDAC2 (p-HDAC2) blot luminescence intensity/total HDAC2 (HDAC2) blot luminescence intensity. Each bar represents the mean of 3 ± SEM or 2 independent measures. A one-way ANOVA with Dunett post-test was used to compare the mean of each test group with the mean of the control group (H2O2 + LPS-treated group). The difference was considered significant for a *p*-value < 0.05. (\*) indicate *p* < 0.05 versus **Figure 8.** HDAC2 phosphorylation induced by H2O<sup>2</sup> + LPS in A549 cells. Effect of NAC and LY294002 (**A**) and the BUD:HPβCD complex versus HPβCD (**B**) after 2 h of incubation. Data are expressed in absolute values (**A**/**B**; with representative blots) or in relative values (in comparison with the pHDAC2/HDAC2 ration of cells incubated with H2O<sup>2</sup> + LPS; **C**/**D**). HDAC2 phosphorylation was quantified after a Western blot by measuring the proportion of phosphorylated-HDAC2 (p-HDAC2) blot luminescence intensity/total HDAC2 (HDAC2) blot luminescence intensity. Each bar represents the mean of 3 ± SEM or 2 independent measures. A one-way ANOVA with Dunett post-test was used to compare the mean of each test group with the mean of the control group (H2O<sup>2</sup> + LPS-treated group). The difference was considered significant for a *p*-value < 0.05. (\*) indicate *p* < 0.05 versus control group.

#### control group. *2.6. BUD:HP*β*CD Complex and HP*β*CD Attenuate H2O<sup>2</sup>* + *LPS-Induced Inflammatory Response in THP-1 Cells*

*2.6. BUD:HPβCD Complex and HPβCD Attenuate H2O2 + LPS-Induced Inflammatory Response in THP-1 Cells*  Since persistent inflammatory response in the lung is a major feature of smoking-induced COPD [31], the anti-inflammatory potential of the BUD:HPβCD complex in comparison with BUD or HPβCD was evaluated. Thus, the effect of the BUD:HPβCD complex, BUD or HPβCD on H2O2 + LPSinduced IL-8 release in A549 cells and TH-P1 cells was determined. Because A549 cells appeared not sensitive to LPS [40], phorbol myristate acetate-activated THP-1 (A-THP-1) cells [41], a widely used Since persistent inflammatory response in the lung is a major feature of smoking-induced COPD [31], the anti-inflammatory potential of the BUD:HPβCD complex in comparison with BUD or HPβCD was evaluated. Thus, the effect of the BUD:HPβCD complex, BUD or HPβCD on H2O<sup>2</sup> + LPS-induced IL-8 release in A549 cells and TH-P1 cells was determined. Because A549 cells appeared not sensitive to LPS [40], phorbol myristate acetate-activated THP-1 (A-THP-1) cells [41], a widely used model for human monocytes, which are highly sensitive to LPS treatment, were used. For the sake of comparison, we also treated A549 cells with TNF-α for inflammatory stress.

model for human monocytes, which are highly sensitive to LPS treatment, were used. For the sake of

comparison, we also treated A549 cells with TNF-α for inflammatory stress.

10).

First, the incubation of A-THP-1 cells with H2O<sup>2</sup> + LPS for 2 h was significantly associated with an 8.6-fold increase in IL-8 release (Figure 9A). Treatment with LPS alone induced a significant 6.6-fold increase, whereas H2O<sup>2</sup> alone did not induce any effects on IL-8 expression [41]. First, the incubation of A-THP-1 cells with H2O2 + LPS for 2 h was significantly associated with an 8.6-fold increase in IL-8 release (Figure 9A). Treatment with LPS alone induced a significant 6.6 fold increase, whereas H2O2 alone did not induce any effects on IL-8 expression [41]. Concomitant incubation of the BUD:HPβCD complex with H2O2 + LPS was associated with a

Concomitant incubation of the BUD:HPβCD complex with H2O<sup>2</sup> + LPS was associated with a lower release of IL-8 of approximately 45% no matter the concentration of the BUD:HPβCD complex used (Figure 9D). In the presence of BUD, a decrease in of IL-8 release was also observed. The effect was similar with that induced by the BUD:HPβCD complex (budesonide 1 and 10 µM), but higher at higher budesonide concentration (100 µM) (67%) (Figure 9B). The presence of HPβCD was also associated with a non-dependent dose-type decrease in IL-8 release. The effect was slightly lower as compared to BUD and the BUD:HPβCD complex (approximately 34%) (Figure 9C). These results suggest that the BUD:HPβCD complex and BUD have similar anti-inflammatory properties, except at high concentrations at which budesonide was more efficient. lower release of IL-8 of approximately 45% no matter the concentration of the BUD:HPβCD complex used (Figure 9D). In the presence of BUD, a decrease in of IL-8 release was also observed. The effect was similar with that induced by the BUD:HPβCD complex (budesonide 1 and 10 µM), but higher at higher budesonide concentration (100 µM) (67%) (Figure 9B). The presence of HPβCD was also associated with a non-dependent dose-type decrease in IL-8 release. The effect was slightly lower as compared to BUD and the BUD:HPβCD complex (approximately 34%) (Figure 9C). These results suggest that the BUD:HPβCD complex and BUD have similar anti-inflammatory properties, except at high concentrations at which budesonide was more efficient.

**Figure 9.** IL-8 release by A-THP-1 cells after treatment with H2O2, LPS or H2O2 + LPS for 2 h (**A**) and effect of BUD (**B**), HPβCD (**C**), and BUD:HPβCD complex (**D**) on H2O2 + LPS-induced IL-8 release. IL-8 release was measured in the extracellular medium by sandwich ELISA. Each bar represents the mean ± SEM of 3 independent means of triplicated measures. A one-way ANOVA with Dunett posttest was used to compare the mean of each test group with the mean of untreated group or H2O2 + LPS-treated group. The difference was considered significant for a *p*-value < 0.05; (aaa), indicate *p* < **Figure 9.** IL-8 release by A-THP-1 cells after treatment with H2O<sup>2</sup> , LPS or H2O<sup>2</sup> + LPS for 2 h (**A**) and effect of BUD (**B**), HPβCD (**C**), and BUD:HPβCD complex (**D**) on H2O<sup>2</sup> + LPS-induced IL-8 release. IL-8 release was measured in the extracellular medium by sandwich ELISA. Each bar represents the mean ± SEM of 3 independent means of triplicated measures. A one-way ANOVA with Dunett post-test was used to compare the mean of each test group with the mean of untreated group or H2O<sup>2</sup> + LPS-treated group. The difference was considered significant for a *p*-value < 0.05; (aaa), indicate *p* < 0.001 versus untreated group; (\*), (\*\*), and (\*\*\*) correspond to *p* < 0.05, 0.01 and 0.001 versus H2O<sup>2</sup> + LPS-treated group, respectively.

LPS-treated group, respectively. A similar effect on IL-8 release was reported when comparing the effect of TNFα on A549 cells A similar effect on IL-8 release was reported when comparing the effect of TNFα on A549 cells with the effect LPS on A-THP1. An anti-inflammatory potential of BUD (1 µM) and the BUD:HPβCD complex (1:25 µM) was observed with a lower potential of HPβCD to decrease IL-8 release (Figure 10).

0.001 versus untreated group; (\*), (\*\*), and (\*\*\*) correspond to *p* < 0.05, 0.01 and 0.001 versus H2O2 +

with the effect LPS on A-THP1. An anti-inflammatory potential of BUD (1 µM) and the BUD:HPβCD complex (1:25 µM) was observed with a lower potential of HPβCD to decrease IL-8 release (Figure

*Molecules* **2020**, *25*, x 12 of 21

**Figure 10.** IL-8 release by A549 cells after treatment with H2O2 + TNF-α for 2 h (**A**) and effect of the BUD:HPβCD complex versus BUD and HPβCD on H2O2 + TNF-α-induced IL-8 release (**B**). IL-8 release was measured in the extracellular medium by sandwich ELISA. Results on panel B were normalized relative to untreated cells (0%) and H2O2 + TNF-α-treated cells (100%). Each bar represents the mean ± SEM of 3 independent means of triplicated measures. (\*), (\*\*), and (\*\*\*), indicate, **Figure 10.** IL-8 release by A549 cells after treatment with H2O<sup>2</sup> + TNF-α for 2 h (**A**) and effect of the BUD:HPβCD complex versus BUD and HPβCD on H2O<sup>2</sup> + TNF-α-induced IL-8 release (**B**). IL-8 release was measured in the extracellular medium by sandwich ELISA. Results on panel B were normalized relative to untreated cells (0%) and H2O<sup>2</sup> + TNF-α-treated cells (100%). Each bar represents the mean ± SEM of 3 independent means of triplicated measures. (\*), (\*\*), and (\*\*\*), indicate, respectively, *p* < 0.05, 0.01, and 0.001 versus non-treated cells (**A**) or H2O2+LPS-treated cells (**B**). BUD:HPβCD complex versus BUD and HPβCD on H2O2 + TNF-α-induced IL-8 release (**B**). IL-8 release was measured in the extracellular medium by sandwich ELISA. Results on panel B were normalized relative to untreated cells (0%) and H2O2 + TNF-α-treated cells (100%). Each bar represents the mean ± SEM of 3 independent means of triplicated measures. (\*), (\*\*), and (\*\*\*), indicate, respectively, *p* < 0.05, 0.01, and 0.001 versus non-treated cells (**A**) or H2O2+LPS-treated cells (**B**).

As HDAC2 is recruited by the activated glucocorticoid receptor to repress the transcription of

respectively, *p* < 0.05, 0.01, and 0.001 versus non-treated cells (**A**) or H2O2+LPS-treated cells (**B**). As HDAC2 is recruited by the activated glucocorticoid receptor to repress the transcription of proinflammatory genes [42] and to study the potential role of HDAC2 in the protection afforded by the BUD:HPβCD complex in comparison with BUD or HPβCD on IL-8 release, IL-8 release induced by TNF-α in conditions where cells were preincubated with or without trichostatin, a pharmacological HDAC2 inhibitor [43], was measured. We pretreated for 30 min A549 cells with increasing concentrations of trichostatin (0–250 nM) and determined IL-8 release after incubation for As HDAC2 is recruited by the activated glucocorticoid receptor to repress the transcription of proinflammatory genes [42] and to study the potential role of HDAC2 in the protection afforded by the BUD:HPβCD complex in comparison with BUD or HPβCD on IL-8 release, IL-8 release induced by TNF-α in conditions where cells were preincubated with or without trichostatin, a pharmacological HDAC2 inhibitor [43], was measured. We pretreated for 30 min A549 cells with increasing concentrations of trichostatin (0–250 nM) and determined IL-8 release after incubation for 2 h of cells with TNF-α (20 ng/mL) and BUD:HPβCD complex or BUD or HPβCD (Figure 11). proinflammatory genes [42] and to study the potential role of HDAC2 in the protection afforded by the BUD:HPβCD complex in comparison with BUD or HPβCD on IL-8 release, IL-8 release induced by TNF-α in conditions where cells were preincubated with or without trichostatin, a pharmacological HDAC2 inhibitor [43], was measured. We pretreated for 30 min A549 cells with increasing concentrations of trichostatin (0–250 nM) and determined IL-8 release after incubation for 2 h of cells with TNF-α (20 ng/mL) and BUD:HPβCD complex or BUD or HPβCD (Figure 11).

2 h of cells with TNF-α (20 ng/mL) and BUD:HPβCD complex or BUD or HPβCD (Figure 11).

**0 50 100 150 200 250 300 Trichostatin [nM] Figure 11.** Percentage of IL-8 released after A549 cells pretreatment for 30 min with trichostatin (TSA) and incubation for 2 h with TNF-α in presence of BUD:HPβCD complex, BUD or HPβCD. Results are **Figure 11.** Percentage of IL-8 released after A549 cells pretreatment for 30 min with trichostatin (TSA) and incubation for 2 h with TNF-α in presence of BUD:HPβCD complex, BUD or HPβCD. Results are expressed in percentage of IL-8 released. 100% corresponds to cells preincubated for 30 min with TSA and incubated for 2 h with TNF-α only. IL-8 release was measured in the extracellular medium by sandwich ELISA. Data are from 3 independent experiments in triplicates. ▼HPβCD; ♦ **Figure 11.** Percentage of IL-8 released after A549 cells pretreatment for 30 min with trichostatin (TSA) and incubation for 2 h with TNF-α in presence of BUD:HPβCD complex, BUD or HPβCD. Results are expressed in percentage of IL-8 released. 100% corresponds to cells preincubated for 30 min with TSA and incubated for 2 h with TNF-α only. IL-8 release was measured in the extracellular medium by sandwich ELISA. Data are from 3 independent experiments in triplicates. HHPβCD; BUD:HPβCD; BUD.

expressed in percentage of IL-8 released. 100% corresponds to cells preincubated for 30 min with TSA

and incubated for 2 h with TNF-α only. IL-8 release was measured in the extracellular medium by sandwich ELISA. Data are from 3 independent experiments in triplicates. ▼HPβCD; ♦ BUD:HPβCD; ■ BUD. IL-8 release induced by TNFα was markedly reduced (around 70%) by BUD and BUD:HPβCD whereas HPβCD alone did not shown any effect or a very slight effect. When trichostatin was used BUD:HPβCD; ■ BUD. IL-8 release induced by TNFα was markedly reduced (around 70%) by BUD and BUD:HPβCD whereas HPβCD alone did not shown any effect or a very slight effect. When trichostatin was used in preincubation to inhibit the HDAC2 activity, the IL-8 release was increased, in a dose-dependent fashion in the presence of BUD or BUD:HPβCD. BUD:HPβCD failed to improve the response of IL-8 release induced by TNFα was markedly reduced (around 70%) by BUD and BUD:HPβCD whereas HPβCD alone did not shown any effect or a very slight effect. When trichostatin was used in preincubation to inhibit the HDAC2 activity, the IL-8 release was increased, in a dose-dependent fashion in the presence of BUD or BUD:HPβCD. BUD:HPβCD failed to improve the response of glucocorticoids in the condition of HDAC2 inhibition. Again, no or a very slight effect was observed with HPβCD (Figure 11).

in preincubation to inhibit the HDAC2 activity, the IL-8 release was increased, in a dose-dependent fashion in the presence of BUD or BUD:HPβCD. BUD:HPβCD failed to improve the response of
