Synthesis and Anticancer Activity of A-Ring-Modified Derivatives of Dihydrobetulin

Multidrug resistance (MDR) is a common phenomenon in clinical oncology, whereby cancer cells become resistant to chemotherapeutic drugs. A common MDR mechanism is the overexpression of ATP-binding cassette efflux transporters in cancer cells, with P-glycoprotein (P-gp) being one of them. New 3,4-seco-lupane triterpenoids, and the products of their intramolecular cyclization with the removed 4,4-gem-dimethyl group, were synthesized by the selective transformations of the A-ring of dihydrobetulin. Among the semi-synthetic derivatives, the MT-assay-enabled methyl ketone 31 (MK), exhibiting the highest cytotoxicity (0.7–16.6 µM) against nine human cancer cell lines, including P-gp overexpressing subclone HBL-100/Dox, is identified. In silico, MK has been classified as a potential P-gp-inhibitor; however, the Rhodamine 123 efflux test, and the combined use of P-gp-inhibitor verapamil with MK in vitro, showed the latter to be neither an inhibitor nor a substrate of P-gp. As the studies have shown, the cytotoxic effect of MK against HBL-100/Dox cells is, arguably, induced through the activation of the ROS-mediated mitochondrial pathway, as evidenced by the positive Annexin V-FITC staining of apoptotic cells, the cell cycle arrest in the G0/G1 phase, mitochondrial dysfunction, cytochrome c release, and the activation of caspase-9 and -3.


Introduction
Over centuries, people have been employing bioactive natural compounds for medicinal purposes, e.g., as part of medicinal plant extracts. Therefore, it is not surprising that, with the development and upgrade of chemical methods (extraction, chromatography, and physicochemical techniques for identifying individual compounds), and within such sections of chemistry as phytochemistry, the chemistry of natural compounds and medicinal chemistry, a great deal of natural compounds and their semisynthetic derivatives began to be actively employed as medicinal products. Thus, over the period 1981-2019, about 25% of approved anticancer drugs are either small molecules of natural origin, or derivations from those [1].
Currently, chemotherapy occupies a leading position in clinical oncology amid the available cancer treatments, such as radiation therapy, surgery, immunotherapy, or these methods combined. Simultaneously, the use of chemotherapy in cancer treatment provokes the development of multidrug resistance (MDR), when tumors become tolerant to multiple anticancer drugs. Multiple MDR mechanisms limit the pharmacological efficacy of cancer treatment and are also responsible for the majority of cancer recurrences and are associated with high rates of cancer mortality [2]. One of the MDR mechanism's developments is related to the overexpression of a key member of the ATP-binding cassette (ABC) family of transporters, P-glycoprotein (P-gp), a transmembrane drug carrier that promotes the efflux of endogenous xenobiotics through cell membranes [2,3]. P-gp is the best characterized Int. J. Mol. Sci. 2023, 24  The ozonolysis of the exomethylene fragment in 3,4-seco-lupane triterpenoids 2, 3 gave rise to methyl ketones 20, 21 which readily underwent intramolecular oxonitrile cyclization in t-BuOH in the presence of t-BuOK to form compounds 22, 23 containing an α,β-unsaturated nitrile fragment in the pentacyclic A-ring (Scheme 2).  We have studied the reactions of methyl ketones 20 and 21 with pyridinium bromide perbromide (C5H5NHBr3) in acetic acid. Methyl ketone 20, with an equimolar amount of C5H5NHBr3 at room temperature, gave a mixture of 24-mono-(28) and 24,24-dibromo-(29) derivatives in 70% and 10% yields, respectively. In turn, methyl ketone 21 formed only 24-monobromo derivative 30 in 76% yield under these conditions (Scheme 3). The reaction of methyl ketone 20 with an excess of C5H5NHBr3 under reflux resulted in the formation of only dibromo derivative 29 (80%), while methyl ketone 21 gave, under analogous conditions, a mixture of di-(31, 60%) and tri-(32, 15%) bromo-substituted methyl ketones (Scheme 4). It should be noted that, in contrast with the oleanane analogs, the elimination of CHBr3 with the formation of 2-cyano-4-oic acids did not occur during the exhaustive bromination of methyl ketones 20, 21, regardless of the molar excess of C5H5NHBr3 and the boiling duration of the reaction mixture.

NC
The structures of the new compounds were confirmed by the 1D NMR ( 1 H-NMR, 13 C-NMR) and MS methods. The determination of the configuration at C(2) and C(3) stereogenic centers in compounds 9 and 11, as well as at C(2) and C(4) in compounds 18 and 19, was carried out using the data of 2D NMR spectra (HMBC, HSQC, NOESY, COZY) of compounds 11 and 19, with the results shown in the Experimental Section and Supplementary Materials.
The structure-activity relationship (SAR) trends are summarized as follows. The structural variations around the C-28 position of the lupane skeleton of 24-bromo-derivatives had a strong influence. This was, probably, the outcome of steric hindrance, on the part of the bulky 28-benzoyl substituent, by the manifestation of cytotoxic activity by compounds 28, 29. The presence of 28-carbomethoxy substituent gave rise to major enhancement of potency, as exemplified by derivatives 30, 31. The degree of cytotoxicity was directly related to the amount of bromine atoms in the 3,4-seco-lupane fragment. High cytotoxic values (3.6-13.5 µM) of 24-monobromethyl ketone 30 were observed relative to the HBL-100/Dox, HBL-100, HCT116, and RD TE32 cells, whereas the emergence of two bromine atoms in the structure of methyl ketone 31 conspicuously heightened cytotoxicity against all the cell lines (0.7-16.6 µM). The 3,4-seco-lupane cyclization products were mostly non-cytotoxic. However, derivatives 12, 14, and 19 showed moderate activity (16.0-77.8 µM) against almost all the tested cancer cell lines, with the relationship between the structure and cytotoxicity of these compounds not being evident.
The selectivity index (SI) for the most active compounds was calculated as a simple ratio of IC 50 values between healthy and cancer cells in order to assess the selective toxic effect of the studied derivatives 12, 14, 19, 30, and 31 against cancer cells and predict their therapeutic potential. As is apparent from Table S2, 24,24-dibromo-substituted methyl ketone 31 (MK) displayed the best SI, especially against HBL-100/Dox, HBL-100, and RD TE32 cells (SI = 6. 39-16.43).
Currently, plant-derived natural products (including triterpenoids and their derivatives) with selective cytotoxicity against cancer cells are also regarded as a worthwhile source of new therapeutic MDR reversal agents characterized by P-gp inhibitors [7,27,28]. The synthesized MK proved to be the most active compound against the P-gp-overexpressing cell line HBL-100/Dox. A detailed analysis of individual molecular properties [29] showed molecular hydrophobicity (MlogP) to be a determining descriptor for identifying P-gp inhibitors. The hydrophobicity, calculated by the Web-tools (MlogP > 2.93 [30]), predicted the P-glycoprotein-binding inhibitory properties as being foreseeable for MK (Table 2).  The interactions between the compound and P-gp leading to MDR reversion may take place due to the competition between substrates for P-gp or the non-competitive inhibition of P-gp activity. The prediction of a possibility for the compounds to act as a substrate or an inhibitor of P-gp, as performed using freely accessible Web-tools ADMETlab 2.0 (https://admetmesh.scbdd.com/ (accessed on 1 March 2023)), pkCSM (http://biosig. unimelb.edu.au/ (accessed on 1 March 2023)), and PgpRules (https://pgprules.cmdm.tw/ (accessed on 1 March 2023)), based on a classification algorithm and regression tree [31], enabled MK and the reference drug verapamil to be classified as the inhibitors of this protein (Table 2). Concurrently, the well-known P-gp inhibitor and MDR reversal agent verapamil were also classified as P-gp substrates.
The ability of MK to prevent the efflux of typical P-gp substrate Rhodamine 123 (Rh123) was evaluated by the flow cytometric analysis of efflux by means of the dye accumulation technique. The Rh123 test is convenient for characterizing the P-gp inhibitory potential of various agents interacting with the Rh123-binding site of P-gp (so-called "Rsite") [32]. As shown in Figure 1, the efflux of Rh123 from HBL-100/Dox cells was active after 60 min of incubation with the tested compound dosed close to the IC 50 value. The efflux values appeared to be significantly different from those observed with verapamil (p < 0.05). The level of MAF value of MK (0.22) was distinctly inferior to those of verapamil (0.93), indicating an extremely low P-gp inhibitory activity of the synthesized compound. To confirm the absence of competitive inhibition, in addition to the efflux test, we performed an experiment aimed at revealing the effects of MK on resistant cells in the presence of verapamil. In case of blocking the P-gp activity, the cytotoxic compound, being a substrate of protein, should have had a more pronounced cytotoxic effect. As in the Rh123 efflux test, the effect of MK was not dependent on the activity of P-gp in the case when the addition of 30 µM verapamil led to the suppression of P-gp-activity; that is, the number of living cells in the compound, and its combination with 30 µM verapamil, remained unchanged (Figure 2a), whereas the IC50 value of Dox against HBL-100/Dox cells had diminished 37-fold (0.8 ± 0.1 µM) with the use of Dox in combi- To confirm the absence of competitive inhibition, in addition to the efflux test, we performed an experiment aimed at revealing the effects of MK on resistant cells in the presence of verapamil. In case of blocking the P-gp activity, the cytotoxic compound, being a substrate of protein, should have had a more pronounced cytotoxic effect. As in the Rh123 efflux test, the effect of MK was not dependent on the activity of P-gp in the case when the addition of 30 µM verapamil led to the suppression of P-gp-activity; that is, the number of living cells in the compound, and its combination with 30 µM verapamil, remained unchanged (Figure 2a), whereas the IC 50 value of Dox against HBL-100/Dox cells had diminished 37-fold (0.8 ± 0.1 µM) with the use of Dox in combination with verapamil ( Figure 3b). Thus, despite the encouraging data provided by the prediction services of web-tools (Table 2), both tests in vitro indicated the cytotoxic effect of MK against P-gp-overexpressing cells as being in no way related to its interaction with the ABC-transporter P-gp.
To confirm the absence of competitive inhibition, in addition to the efflux test, we performed an experiment aimed at revealing the effects of MK on resistant cells in the presence of verapamil. In case of blocking the P-gp activity, the cytotoxic compound, being a substrate of protein, should have had a more pronounced cytotoxic effect. As in the Rh123 efflux test, the effect of MK was not dependent on the activity of P-gp in the case when the addition of 30 µM verapamil led to the suppression of P-gp-activity; that is, the number of living cells in the compound, and its combination with 30 µM verapamil, remained unchanged (Figure 2a), whereas the IC50 value of Dox against HBL-100/Dox cells had diminished 37-fold (0.8 ± 0.1 µM) with the use of Dox in combination with verapamil ( Figure 3b). Thus, despite the encouraging data provided by the prediction services of web-tools (Table 2), both tests in vitro indicated the cytotoxic effect of MK against P-gp-overexpressing cells as being in no way related to its interaction with the ABC-transporter P-gp. While evaluating the effect of the combination of Dox and MK, combined at subtoxic concentration (0.6 µM), against resistant cells (Figure 2c), no potentiation in the effectiveness of Dox: 29.6 ± 0.9 µM vs. 30.1 ± 0.5 µM (taken alone vs. in combination, p > 0.05) was observed. Thus, any other interaction between the triterpenoid MK and the chemotherapeutic drug differing from P-gp inhibition was also excluded.  Since a continuous MK treatment for 12-24 h led to a significant decrease in cell viability, the DNA content of HBL-100 and HBL-100/Dox cells in sub-G0/G1 phase, G0/G1 phase, S phase, and G2/M phase was determined using the flow cytometry technique within the same time interval. As presented in Figure 4a, the HBL-100 cell fractions in sub-G0/G1 phase (indicative of cell death) increased from 1.44 ± 0.44 to 8.22 ± 0.59%, with marked differences (p < 0.05) recorded after 16 h exposure to MK. After 24 h incubation, While evaluating the effect of the combination of Dox and MK, combined at subtoxic concentration (0.6 µM), against resistant cells (Figure 2c), no potentiation in the effectiveness of Dox: 29.6 ± 0.9 µM vs. 30.1 ± 0.5 µM (taken alone vs. in combination, p > 0.05) was ob-served. Thus, any other interaction between the triterpenoid MK and the chemotherapeutic drug differing from P-gp inhibition was also excluded.
Then, MK was investigated more thoroughly against HBL-100 and HBL-100/Dox cells to prove a common mechanism of cytotoxic action. At first, we analyzed the changes in HBL-100 ( Figure 3a) and HBL-100/Dox (Figure 3b) cell proliferation after being exposed to MK in a time-course manner (6, 12, 18, 24, 48, 72 h) by MTT assay. Exposure to MK at doses corresponding to IC 50 values induced a time-dependent reduction in cell proliferation in comparison to the untreated cells. As apparent from Figure 3, the cell viability curves began to descend after 12 h incubation. After 18 h treatment, the most significant drop in HBL-100 and HBL-100/Dox viability was observed, with a 50% death of the cells attained after 72 h exposure.
Since a continuous MK treatment for 12-24 h led to a significant decrease in cell viability, the DNA content of HBL-100 and HBL-100/Dox cells in sub-G0/G1 phase, G0/G1 phase, S phase, and G2/M phase was determined using the flow cytometry technique within the same time interval. As presented in Figure 4a, the HBL-100 cell fractions in sub-G0/G1 phase (indicative of cell death) increased from 1.44 ± 0.44 to 8.22 ± 0.59%, with marked differences (p < 0.05) recorded after 16 h exposure to MK. After 24 h incubation, the proportion of MK-treated cells in the G0/G1 phase increased to 66.48 ± 1.33%, while the proportion of cells in the G2/M phase reduced from 16.66 ± 0.15 to 4.43 ± 0.50% in comparison to the untreated cells. As for the HBL-100/Dox cells (Figure 4b), the sub-G0/G1 cell population increased in a time-dependent manner from 4.84 ± 1.48 to 13.47 ± 2.44% (registered in the 12-24 h interval). This is in accord with a conspicuous reduction in the cells in the S phase from 30.69 ± 0.99 to 8.99 ± 0.42% (from 12 to 24 h). In comparison to the untreated HBL-100/Dox cells (40.34 ± 0.57%), an accumulation in the G0/G1 phase of 52.16 ± 1.94% was observed in MK-treated cells after 17 h of incubation. After 24 h, along with the accumulation of the MK-treated HBL-100/Dox cells in phase G0/G1 (64.14 ± 0.55%), an increase in the G2/M phase (26.87 ± 1.13%), in comparison to the untreated cells (G0/G1 phase: 51.24 ± 1.17%, G2/M phase: 13.19 ± 1.07%), was observed. Therefore, there was no dominant interference with the cell cycle regulation caused by MK; the triterpenoid blocked or slowed down the cell cycle progression through the G0/G1 phase for parental and resistant cells (Tables S3 and S4). Meanwhile, Dox induced the S phase arrest of the HBL-100 and HBL-100/Dox cells, starting at 12 h, and without any significant change in the sub-G0/G1 phase.
The most significant changes in sub-G0/G1 population (p < 0.05) for the HBL-100 and HBL-100/Dox cells were recorded after 16 h, following MK treatment. With fractional DNA content being a specific marker of cell death most often attributed to the induction of apoptosis in the case of triterpenoids [11,12], we additionally evaluated the dose-response apoptotic MK effect in the HBL-100 and HBL-100/Dox cells by Annexin V-FITC detection and propidium iodide (PI) staining using the flow cytometry technique. Annexin V has a strong binding affinity with phosphatidylserine (PS), a membrane phospholipid that moves from the inside of the cell membrane to its outer side during apoptosis, while PI has the ability to bind to DNA and can only penetrate necrotic or late apoptotic cells [33]. Therefore, the proportions of dead (Annexin V−/PI+) to viable (Annexin V−/PI−) cells, as well as early (EA, Annexin V+/PI−) to late (LA, Annexin V+/PI+) apoptotic cells, were measured for quantitative comparison (Table S5). In the experiment, the HBL-100 and HBL-100/Dox cells were treated with different doses of MK (IC 50  The most significant changes in sub-G0/G1 population (p < 0.05) for the HBL-100 and HBL-100/Dox cells were recorded after 16 h, following MK treatment. With fractional DNA content being a specific marker of cell death most often attributed to the induction of apoptosis in the case of triterpenoids [11,12], we additionally evaluated the dose-response apoptotic MK effect in the HBL-100 and HBL-100/Dox cells by Annexin V-FITC detection and propidium iodide (PI) staining using the flow cytometry technique. Annexin V has a strong binding affinity with phosphatidylserine (PS), a membrane phospholipid that moves from the inside of the cell membrane to its outer side during apoptosis, while PI has the ability to bind to DNA and can only penetrate necrotic or late apoptotic cells [33]. Therefore, the proportions of dead (Annexin V−/PI+) to viable (Annexin V−/PI−) cells, as well as early (EA, Annexin V+/PI−) to late (LA, Annexin V+/PI+) apoptotic cells, were measured for quantitative comparison (Table S5)  The double-fluorescent staining was employed to further validate the underlying mechanisms of apoptosis. The HBL-100 and HBL-100/Dox cells were stained with Hoechst 33342 and TMRE (tetramethylrhodamine ethyl ester perchlorate) after being treated with 1.0 and 1.8 µM MK for 16 h, respectively. Hoechst 33342 is a blue fluorescent dye that binds itself to DNA and it can be used to observe nuclear condensation by fluorescence microscopy. In healthy, untreated cells, nuclei appear round and evenly stained. In apoptotic cells, nuclei are generally fragmented and stained more intensely because of the condensation of DNA [34]. TMRE is a cationic red-orange fluorescent dye used for measuring mitochondrial membrane potential (MMP). The depolarized or inactive mitochondria decrease MMP and fail to sequester TMRE, so that apoptotic cells exhibit a weak fluorescence, while the accumulation of TMRE indicate the liveliness of cells [35]. ± 0.82 to 10.37 ± 1.01%. There were no statistically significant differences in the case of the EA HBL-100/Dox cells (5.04 ± 0.51 vs. 5.76 ± 0.38%, p > 0.05) found; by contrast, the proportion of LA cells increased from 4.74 ± 0.44 to 10.94 ± 0.81% after MK treatment. These results provided evidence that MK provokes a notable increase in cellular apoptosis in a dose-dependent manner. In the case of Dox, the population of EA cells increased to 12.35 ± 0.42 (HBL-100 cells) and 2.56 ± 0.36% (HBL-100/Dox cells) in comparison with the untreated cells (2.12 ± 0.20 and 1.41 ± 0.25%, respectively). The double-fluorescent staining was employed to further validate the underlying mechanisms of apoptosis. The HBL-100 and HBL-100/Dox cells were stained with Hoechst 33342 and TMRE (tetramethylrhodamine ethyl ester perchlorate) after being treated with 1.0 and 1.8 µM MK for 16 h, respectively. Hoechst 33342 is a blue fluorescent dye that binds itself to DNA and it can be used to observe nuclear condensation by fluorescence microscopy. In healthy, untreated cells, nuclei appear round and evenly stained. In apoptotic cells, nuclei are generally fragmented and stained more intensely because of the condensation of DNA [34]. TMRE is a cationic red-orange fluorescent dye used for measuring mitochondrial membrane potential (MMP). The depolarized or inactive mitochondria decrease MMP and fail to sequester TMRE, so that apoptotic cells ex- Our observation showed the untreated HBL-100 and HBL-100/Dox cells to demonstrate weak blue Hoechst 33342 fluorescence whereas, in the groups treated with MK and Dox, the apoptotic cells exhibited typical nuclear changes, including bright staining and condensed or fragmented nuclei (Figure 6a,b). Moreover, the number of cells with nuclear morphology changes increased after MK treatment to 21.2% (HBL-100 cells) and 33.1% (HBL-100/Dox cells) in comparison with the untreated controls (HBL-100: 3.1%; HBL-100/Dox: 4.1%), while the number of Dox-treated cells was 25.2% (HBL-100 cells) and 9.9% (HBL-100/Dox cells). A significant (p < 0.05) reduction in MMP, as indicated by low TMRE fluorescence intensity in the MK-treated HBL-100 and HBL-100/Dox cells, when compared with the untreated control (Figure 6c,d), was observed.
The flow cytometry and fluorescent staining data, analyzed comparatively, indicated a more noticeable triterpenoid treatment effect 16 h later, when observed against HBL-100/Dox cells and compared with HBL-100 cells. Thus, an increased nuclear density (or karyopycnosis) was visualized in parental cells (Figure 6a), while the subsequent stage of nuclear fragmentation (or karyorrhexis) was recorded in resistant cells (Figure 6b). Furthermore, in comparison with the untreated cells, a significant increase in the fractional DNA content was recorded much earlier in HBL-100/Dox cells in comparison to the HBL-100 cells treated with MK for 12 or 16 h, respectively (Figure 4). This observation is in concordance with the Annexin V-FITC/PI cellular staining after 16 h of MK treatment, when the apoptosis-associated externalization of PS and the loss of plasma membrane integrity were more pronounced in the HBL-100/Dox cells in comparison to the HBL-100 cells ( Figure 5). Thus, the cells overexpressing P-gp appeared to be significantly more susceptible to the MK's toxic effect. strate weak blue Hoechst 33342 fluorescence whereas, in the groups treated with MK and Dox, the apoptotic cells exhibited typical nuclear changes, including bright staining and condensed or fragmented nuclei (Figure 6a,b). Moreover, the number of cells with nuclear morphology changes increased after MK treatment to 21.2% (HBL-100 cells) and 33.1% (HBL-100/Dox cells) in comparison with the untreated controls (HBL-100: 3.1%; HBL-100/Dox: 4.1%), while the number of Dox-treated cells was 25.2% (HBL-100 cells) and 9.9% (HBL-100/Dox cells). A significant (p < 0.05) reduction in MMP, as indicated by low TMRE fluorescence intensity in the MK-treated HBL-100 and HBL-100/Dox cells, when compared with the untreated control (Figure 6c,d), was observed. Mitochondrial metabolic plasticity is known to contribute to the drug resistance of cancer cells [36]. One of the mechanisms of MDR is an increased antioxidant system that protects cells from the increased oxidative stress caused by chemotherapy, which might be attained by reducing mitochondrial biogenesis. For example, sensitive and Dox-resistant osteosarcoma cells have a different mitochondrial profile, in particular, demonstrating a reduced mitochondrial membrane mass, potential, and activity in comparison to Dox-sensitive cells [37]. As apparent from Figure 6c, the HBL-100 and HBL-100/Dox cells also had different basal levels of mitochondrial activity based on TMRE staining, demonstrating a reduction in MMP in native Dox-resistant cells. After MK treatment, a noticeable decrease in MMP in the both cell lines, which indicated the participation of mitochondria in MKmediated apoptosis (Figure 6), was observed.
Mitochondrial membrane permeability is a crucial step in the apoptotic pathway cascades. In the mitochondrial apoptotic pathway, the destruction of the outer mitochondrial membrane leads to the release of cytochrome c from mitochondria into cytosol. Initiator caspase-9 is activated earlier than effector caspase-3 by binding to a complex containing Apaf-1 and cytochrome c. When caspase-9 proteolytically cleaves downstream effector caspase-3, the cleaved caspase-3 transfers to the nucleus and initiates the degradation phase of apoptosis, including DNA fragmentation, by involving the DNA repair enzymes, such as PARP and ICAD [38]. Betulinic acid, and its numerous semi-synthetic derivatives, most often induce apoptosis in cancer cells through the mitochondrial signaling cascade, which involves the expression of caspase-9 and -3 [11]. We additionally assessed whether the caspase-9 activation leads to the downstream activation of caspase-3, using the flow cytometry technique. We also observed the caspase-3 activation in the HBL-100/Dox cells to be doubly pronounced, in comparison to the HBL-100 cells, after 24 h of MK treatment, with the effect being dose-dependent ( Figure S5). This was in accordance with a marked increase in fragmented DNA in the HBL-100/Dox cells in same time interval (peak sub-G0/G1, Figure 4).
One of the promising strategies of anticancer drug development is the design and development of mitochondria-targeted compounds [36,39]. Among apoptosis-inducing pentacyclic triterpenoids, the lupane triterpenoids with cation moiety can accumulate in the mitochondrial matrix against their concentration gradient, due to the negative potential of the mitochondrial inner membrane, thereby promoting cytochrome c release into the cytoplasm and initiating the mitochondrial pathway of apoptosis by activating the caspase cascade [40,41]. At the same time, hydrophobic betulin and betulinic acid, the parent compounds of MK, directly affect the bioenergetics and membrane behavior of mitochondria isolated from the liver of Wistar rats [42]. MK may act in a similar way, but we could not exclude the modulation of apoptotic cascade due to oxidative stress by production of an abnormally high level of intracellular reactive oxygen species (ROS), which is also characteristic of triterpenoids [43,44]. Therefore, the ROS status was evaluated in the HBL-100/Dox cells after 4 h of exposure to MK, when a decrease in cell viability was not yet observed (Figure 3b), because apoptotic signs were only registered starting at 12 h (Figure 4b). As determined by the H2DCFDA fluorescent staining ( Figure   Figure 7. Western blot analysis of the expression of cytochrome c (a) and the activation of caspase-9 and -8 (b) in HBL-100 and HBL-100/Dox cells after MK treatment for 16 h. β-Actin was used as loading control.
We additionally assessed whether the caspase-9 activation leads to the downstream activation of caspase-3, using the flow cytometry technique. We also observed the caspase-3 activation in the HBL-100/Dox cells to be doubly pronounced, in comparison to the HBL-100 cells, after 24 h of MK treatment, with the effect being dose-dependent ( Figure S5). This was in accordance with a marked increase in fragmented DNA in the HBL-100/Dox cells in same time interval (peak sub-G0/G1, Figure 4).
One of the promising strategies of anticancer drug development is the design and development of mitochondria-targeted compounds [36,39]. Among apoptosis-inducing pentacyclic triterpenoids, the lupane triterpenoids with cation moiety can accumulate in the mitochondrial matrix against their concentration gradient, due to the negative potential of the mitochondrial inner membrane, thereby promoting cytochrome c release into the cytoplasm and initiating the mitochondrial pathway of apoptosis by activating the caspase cascade [40,41]. At the same time, hydrophobic betulin and betulinic acid, the parent compounds of MK, directly affect the bioenergetics and membrane behavior of mitochondria isolated from the liver of Wistar rats [42]. MK may act in a similar way, but we could not exclude the modulation of apoptotic cascade due to oxidative stress by production of an abnormally high level of intracellular reactive oxygen species (ROS), which is also characteristic of triterpenoids [43,44]. Therefore, the ROS status was evaluated in the HBL-100/Dox cells after 4 h of exposure to MK, when a decrease in cell viability was not yet observed (Figure 3b), because apoptotic signs were only registered starting at 12 h ( Figure 4b The results obtained are in good agreement with the supposition that ROS generation by triterpenoids leads to changes in the inner mitochondrial membrane, causing mitochondrial permeability and damaging its integrity. Consequently, the loss of MMP and the release of proapoptotic proteins into the cytosol, followed by the activation of the caspase cascade [43,44], is present.

Chemistry
All the commercially available reagents were purchased from Merck (Darmstadt, Germany), Sigma-Aldrich Pty Ltd. (an affiliate of Merck KGaA, Darmstadt, Germany, or Acros Organics (Geel, Belgium), and used without further purification. Column chromatography was carried out on a Macherey-Nagel 60 Silica (Düren, Germany) (0.063-0.2 The results obtained are in good agreement with the supposition that ROS generation by triterpenoids leads to changes in the inner mitochondrial membrane, causing mitochondrial permeability and damaging its integrity. Consequently, the loss of MMP and the release of proapoptotic proteins into the cytosol, followed by the activation of the caspase cascade [43,44], is present.

Chemistry
All the commercially available reagents were purchased from Merck (Darmstadt, Germany), Sigma-Aldrich Pty Ltd. (an affiliate of Merck KGaA, Darmstadt, Germany, or Acros Organics (Geel, Belgium), and used without further purification. Column chromatography was carried out on a Macherey-Nagel 60 Silica (Düren, Germany) (0.063-0.2 mm). The Sorbfil plates (IMID, Krasnodar, Russia) were employed for thin-layer chromatography (TLC). Melting points were determined on an OptiMelt MPA100 device (Stanford Research Systems, Sunnyvale, CA, USA) at heating rate 1 • C/min. IR spectra (ν, cm −1 ) were recorded on a Bruker IFS 66/S FT-IR spectrometer (Bruker Optik GmbH, Ettlingen, Germany) using a thin film obtained by evaporation from the solution of the substance in CHCl 3 . The 1D spectra ( 1 H NMR and 13 C NMR) and 2D spectra ( 1 H-1 H COSY, 1 H-1 H NOESY, 1 H-13 C HSQC, 1 H-13 C HMBC) were recorded on a Bruker AVANCE II spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany) at 400 MHz ( 1 H NMR) and 100 MHz ( 13 C NMR) in CDCl 3 with hexamethyldisilane as an internal standard. Chemical shifts (δ) and coupling constants (J) were expressed in parts per million (ppm) and Hertz (Hz), respectively. Optical rotation was measured on a Perkine Elmer 341 polarimeter (Perkin Elmer, Waltham, MA, USA) using the sodium D line (589 nm) as a light source for CHCl 3 solutions. Mass spectra (MS) were determined on an Agilent 6890N/5975B chromatograph (Agilent Technologies, Wilmington, NC, USA) equipped with an HP-5 ms UI capillary column (4 m × 0.25 mm, 0.25 µm; 70 eV electron impact). Ozonation reaction was carried out with use of a laboratory ozonizer OGVK-01 (CJSC MELP, St. Petersburg, Russia). Compounds 2, 3 were prepared according to standard procedures [23,24] 13  An amount of 6 Mmol of NBS was added to a solution of 3 mmol compound 2 (or 3) in 50 mL of anhydrous CCl 4 . The mixture was boiled under stirring for 2-4 h. The reaction was monitored by TLC. The solvent was distilled under vacuum, and compounds 6, 7 were purified by column chromatography with SiO 2 using the mixture of petroleum ether-ethylacetate-chloroform (20:1:1) as eluent.

MTT Cell Viability Assay
To investigate the cytotoxic activity of the compounds under study, the MTT colorimetric assay [45] was performed. Briefly, 1 × 10 4 cells suspended in 100 µL of growth medium were seeded in 96-well plates (SPL Life Sciences, Pocheon-si, Republic of Korea), and incubated for 24 h at +37 • C and in 5% CO 2 . All the compounds under study, dissolved in DMSO (PanEco, Moscow, Russia) to concentration 1 × 10 −2 M in advance and further diluted in growth medium, were added to the cells at concentrations ranging from 100 to 0.3125 µM. The DMSO-treated cells (with 1% DMSO, no effect on cell growth) were used as a control, while Dox was used as a reference drug. After 72 h, 20 µL of 5 mg/mL solution of MTT (PanEco, Moscow, Russia), in sterile phosphate-buffered saline (PBS; PanEco, Moscow, Russia), was added to each well and further incubated for 4 h. Then, the medium with the compounds was removed, and the formed formazan crystals were dissolved in 100 µL of DMSO. The optical density of the content of each of the 96-well plates at 544 nm was measured using a microplate reader FLUOstar Optima (BMG Labtech, Ortenberg, Germany). The IC 50 values were determined from dose-dependent curves, using Prism 6.0 (GraphPad Software, Boston, MA, USA).
The time-dependent viability of the HBL-100 and HBL-100/Dox cells exposed to MK was evaluated by MTT assay, as described above. The cells were treated with MK (IC 50 values) for 6, 12, 18, 24, 48, and 72 h. The percentage of the viable cells was calculated according to the following equation: cell viability (%) = 100 × (mean absorbance in the treatment group/mean absorbance in the control group). The results were expressed pictorially using Prism 6.0 (GraphPad Software, Boston, MA, USA).

In Silico P-gp Substrate and P-gp Inhibitor Prediction
Three online services were used to predict whether triterpenoid MK can behave as a substrate or an inhibitor of P-gp. ADMETlab 2.0 (https://admetmesh.scbdd.com/ (accessed on 1 March 2023)), an extended version of the widely used ADMETlab, was used to predict the ADMET profiles for chemicals [46]. The approach, called pkCSM, also provides a platform for predicting and optimizing pharmacokinetic and toxicity properties and relies on distance-based graph signatures. The pkCSM is provided online at http://biosig.unimelb.edu.au/pkcsm/ (accessed on 1 March 2023) [47].
PgpRules is a server providing two separate prediction services, one for P-gp substrate and another one for P-gp inhibitor, with classification and regression trees (CART) treated with MK (at IC 50 and 2 × IC 50 concentrations) for another 24 h period. The cells were detached with accutase (HiMedia Laboratories, Maharashtra, India) and then fixed with 70% ice cold ethanol followed by overnight incubation at 4 • C. The cells were washed twice with PBS, permeabilized using 0.1% Triton X-100 in PBS solution, and subsequently stained with 5 µL of the BD Horizon™ BV605 Rabbit Anti-Active Caspase-3 antibody (BD Biosciences, Franklin Lakes, NJ, USA). The cells were then washed, resuspended in PBS, and analyzed using a CytoFlex S flow cytometer (Beckman Coulter, Brea, CA, USA). Cell fluorescence was evaluated on a CytoFlex S flow cytometer (Beckman Coulter, Brea, CA, USA). The results were analyzed using FlowJo software (ver. X 10.0.7r2, FlowJo Software, San Jose, CA, USA) and represented as geometric mean fluorescence intensity (gMFI).

Statistics
All the experiments were performed in triplicate. The results were presented as mean ± SD values. The differences were considered statistically significant at p < 0.05. For statistical analysis and graph design, GraphPad Prism 6.0 was used. The statistical significance was evaluated by the unpaired t-test and Mann-Whitney test.

Conclusions
Overall, new dihydrobetulin derivatives, including A-seco-derivatives and "triterpenoidsteroid" hybrids, were synthesized. Screening studies of cytotoxic activity revealed 24,24dibromo-substituted methyl ketone 31 (MK) as being the most active anticancer agent, the activity of which was also preserved against the P-gp-overexpressing cancer cells. The studies of interaction between MK and P-gp have shown that MK is not a substrate or an inhibitor of the ABC-transporter P-gp. Regardless of the P-gp status of cancer cells, MK causes apoptotic cell death by increasing ROS production, thereby triggering the depolarization of the inner mitochondrial membrane potential, the intrinsic apoptotic pathway with cytochrome c release, and the activation of caspases-9 and -3.