Anticancer Cytotoxic Activity of Bispidine Derivatives Associated with the Increasing Catabolism of Polyamines

Polyamine (PA) catabolism is often reduced in cancer cells. The activation of this metabolic pathway produces cytotoxic substances that might cause apoptosis in cancer cells. Chemical compounds able to restore the level of PA catabolism in tumors could become potential antineoplastic agents. The search for activators of PA catabolism among bicyclononan-9-ones is a promising strategy for drug development. The aim of the study was to evaluate the biological activity of new 3,7-diazabicyclo[3.3.1]nonan-9-one derivatives that have antiproliferative properties by accelerating PA catabolism. Eight bispidine derivatives were synthetized and demonstrated the ability to activate PA catabolism in regenerating rat liver homogenates. However, only three of them demonstrated a potent ability to decrease the viability of cancer cells in the MTT assay. Compounds 4c and 4e could induce apoptosis more effectively in cancer HepG2 cells rather than in normal WI-38 fibroblasts. The lead compound 4e could significantly enhance cancer cell death, but not the death of normal cells if PAs were added to the cell culture media. Thus, the bispidine derivative 4e 3-(3-methoxypropyl)-7-[3-(1H-piperazin-1-yl)ethyl]-3,7-diazabicyclo[3.3.1]nonane could become a potential anticancer drug substance whose mechanism relies on the induction of PA catabolism in cancer cells.


Introduction
Polyamines (PAs) are ubiquitous aliphatic polycations (putrescine (Put), spermidine (Spd), and spermine (Spm)) that can interact with biological polyanionic macromolecules [1][2][3][4]. The binding of PA to proteins results in the modulation of the activity of different enzymes [5] and ion channels [6] and supports the functions of cell membranes. Due to their ability to interact with nucleic acids, PAs can influence gene transcription [7,8] and mRNA translation [9,10]. Thus, PAs are essential for different cellular functions, including cell growth and proliferation [11]. In addition to regulating functions in normal cells,
Column chromatography (on III activity alumina, the eluent is benzene:dioxane 5:1) was used for purification of novel bicyclic ketones, nonanes, bicyclic oximes and O-benzoyl oximes. The completion of the reactions was monitored by TLC. 1H and 13C NMR spectroscopies were used to determine the structures of the synthesized substances.
Column chromatography (on III activity alumina, the eluent is benzene:dioxane 5:1) was used for purification of novel bicyclic ketones, nonanes, bicyclic oximes and O-benzoyl oximes. The completion of the reactions was monitored by TLC. 1H and 13C NMR spectroscopies were used to determine the structures of the synthesized substances.
The formation of bicyclic amines was evidenced by the absence of the absorption band of the carbonyl group in the IR spectra of compounds 3a-3e.
When comparing the 13C NMR spectra of 3,7-diazabicyclo[3.3.1]nonanes (3a-3e) with the spectra of the starting bicyclic ketones (2a-2e), it can be seen that they lack the signal of the carbon atom characteristic of the carbonyl group, while in the upfield part of the spectrum, a triplet signal of the carbon atom of the methylene group appears in the 9th position.
The absorption bands of the C=N bond and the OH group are identified in the IR spectra of oximes (5a-5c). The most intense band in 6a-6c corresponds to the benzoyloxycarbonyl group. The 13C NMR spectra of oximes (5a-5c) and their esters (6a-6c) showed resonances for C-9. In the 1H NMR spectra, H1,5 protons were observed as separate The formation of bicyclic amines was evidenced by the absence of the absorption band of the carbonyl group in the IR spectra of compounds 3a-3e.
When comparing the 13C NMR spectra of 3,7-diazabicyclo[3.3.1]nonanes (3a-3e) with the spectra of the starting bicyclic ketones (2a-2e), it can be seen that they lack the signal of the carbon atom characteristic of the carbonyl group, while in the upfield part of the spectrum, a triplet signal of the carbon atom of the methylene group appears in the 9th position.
The absorption bands of the C=N bond and the OH group are identified in the IR spectra of oximes (5a-5c). The most intense band in 6a-6c corresponds to the benzoyloxycarbonyl group. The 13C NMR spectra of oximes (5a-5c) and their esters (6a-6c) showed resonances for C-9. In the 1H NMR spectra, H1,5 protons were observed as separate signals. The weakest signal was assigned to the proton of the NOH group (5a-5c). Signals for the protons of the phenyl ring were observed in the spectrum of the O-benzoyloxime derivatives (6a-6c).
The core bispidine structure of 3,7-diazabicyclo[3.3.1]nonane contains two heterocyclic nitrogens-N3 and N7 ( Figure 1B). Different combinations of side N3-and N7-radicals provide a set of five compounds 4a-4e, three of which were supplied with benzoyloxim partners 7a-7c and were also included in the investigation.

The Activation of Polyamine Catabolism in Regenerating Rat Liver Homogenates
Eight synthetized bispidine derivatives were tested for their ability to activate oxidation of PAs in the homogenates of regenerating rat liver. Put is normally oxidized by diamine oxidase, which is a copper-containing enzyme that produces ammonia and H 2 O 2 [37]. Spd and Spm are substrates of Spd/Spm-acetyltransferase and acetyl-polyamine oxidase. The results demonstrate that all the compounds could activate the oxidation of all three PAs (Put, Spd, Spm) at varying rates ( Figure 2).
The core bispidine structure of 3,7-diazabicyclo[3.3.1]nonane contains two heterocyclic nitrogens-N3 and N7 ( Figure 1B). Different combinations of side N3-and N7-radicals provide a set of five compounds 4a-4e, three of which were supplied with benzoyloxim partners 7a-7c and were also included in the investigation.

The Activation of Polyamine Catabolism in Regenerating Rat Liver Homogenates
Eight synthetized bispidine derivatives were tested for their ability to activate oxidation of PAs in the homogenates of regenerating rat liver. Put is normally oxidized by diamine oxidase, which is a copper-containing enzyme that produces ammonia and H2O2 [37]. Spd and Spm are substrates of Spd/Spm-acetyltransferase and acetyl-polyamine oxidase. The results demonstrate that all the compounds could activate the oxidation of all three PAs (Put, Spd, Spm) at varying rates ( Figure 2). Figure 2. The influence of the tested compounds on the rate of PA oxidation. Rat liver homogenates were incubated with 20 µM compounds, and polyamine oxidase activity was determined for (A) Put, (B) Spd and (C) Spm as described in the experimental section. N = 4. * p ≤ 0.05 vs. control. # p ≤ 0.05 between bicyclononane and its benzoyloxym-free partner. Con., control.
Benzoyloxym compounds might be considered prodrugs with partly blocked activity due to the presence of this specific group that might be eliminated by hepatic enzymes. Our set of bispidine derivatives includes three such "drug-prodrug" pairs: 4a/7a, 4b/7b, and 4c/7c. Pairs 4a/7a and 4c/7c contain a hydrophobic fragment at N3, while the 4b/7b pair contains a terminal OH group at N3 that provides some hydrophilicity to this fragment. The general tendency for oxidation of different PAs against the background of the action of the compounds is the following: only pairs 4b/7b demonstrated true drug-prodrug-like activity, as the accelerating effect on PA catabolism was significantly lowered by benzoyloxymation. Compounds 7a and 7c, by contrast, were more active than their benzoyloxym-free partners 4a and 4a, respectively, which demonstrates that the benzoyloxym moiety might enhance the activatory action of these diazabicyclononanes toward PA oxidation. Compounds 4d and 4e are both benzoyloxym-free but differ by the side radical at N3, namely, 4d contains an aryl moiety, while 4e contains an aliphatic heterocycle. The comparison of these two compounds reveals that the aromatic radical provides stronger PA catabolism activatory properties to the molecule as the PA oxidation becomes Figure 2. The influence of the tested compounds on the rate of PA oxidation. Rat liver homogenates were incubated with 20 µM compounds, and polyamine oxidase activity was determined for (A) Put, (B) Spd and (C) Spm as described in the experimental section. N = 4. * p ≤ 0.05 vs. control. # p ≤ 0.05 between bicyclononane and its benzoyloxym-free partner. Con., control.
Benzoyloxym compounds might be considered prodrugs with partly blocked activity due to the presence of this specific group that might be eliminated by hepatic enzymes. Our set of bispidine derivatives includes three such "drug-prodrug" pairs: 4a/7a, 4b/7b, and 4c/7c. Pairs 4a/7a and 4c/7c contain a hydrophobic fragment at N3, while the 4b/7b pair contains a terminal OH group at N3 that provides some hydrophilicity to this fragment. The general tendency for oxidation of different PAs against the background of the action of the compounds is the following: only pairs 4b/7b demonstrated true drugprodrug-like activity, as the accelerating effect on PA catabolism was significantly lowered by benzoyloxymation. Compounds 7a and 7c, by contrast, were more active than their benzoyloxym-free partners 4a and 4c, respectively, which demonstrates that the benzoyloxym moiety might enhance the activatory action of these diazabicyclononanes toward PA oxidation. Compounds 4d and 4e are both benzoyloxym-free but differ by the side radical at N3, namely, 4d contains an aryl moiety, while 4e contains an aliphatic heterocycle. The comparison of these two compounds reveals that the aromatic radical provides stronger PA catabolism activatory properties to the molecule as the PA oxidation becomes 3-5-fold higher under the action of 4d than that of 4e. Thus, 4d becomes a leader among all the tested compounds in regard to the activation of Spm oxidation.

Toxicity of Bicyclononane Derivatives toward Cancer Cells
To test the cytotoxic activity (loss of cell viability) of the compounds, HepG2 human liver carcinoma cells and WI-38 normal fibroblasts were cultivated in the presence of different concentrations of synthesized compounds, and cell viability and apoptosis induction were measured after 72 h of incubation. Table 1 represents IC50 values toward the chosen cell lines. Only three compounds (4c, 4e, 7a) demonstrated selectivity toward cancer HepG2 cells ( Figure 3A) with significantly less cytotoxicity and higher IC50 toward normal cells ( Figure 3B). Compound 4e was the most active and could induce a gradual decrease in HepG2 cell viability in the range of concentrations of 3-9 µM. Almost all cells were not viable after incubation at higher concentrations. Compound 7a was less active and could induce a gradual decrease in cell viability at concentrations of 7-25 µM. Compound 4c demonstrated moderate activity and gradually suppressed cell growth at concentrations of 3-25 µM. Compound 7a demonstrated the ability to increase the viability of normal WI-38 fibroblasts at low concentrations of 1-9 µM and was therefore excluded from further studies. chosen cell lines. Only three compounds (4c, 4e, 7a) demonstrated selectivity toward cancer HepG2 cells ( Figure 3A) with significantly less cytotoxicity and higher IC50 toward normal cells ( Figure 3В). Compound 4e was the most active and could induce a gradual decrease in HepG2 cell viability in the range of concentrations of 3-9 µM. Almost all cells were not viable after incubation at higher concentrations. Compound 7a was less active and could induce a gradual decrease in cell viability at concentrations of 7-25 µM. Compound 4c demonstrated moderate activity and gradually suppressed cell growth at concentrations of 3-25 µM. Compound 7a demonstrated the ability to increase the viability of normal WI-38 fibroblasts at low concentrations of 1-9 µM and was therefore excluded from further studies.   To investigate the type of cell death, cells were incubated with 25 µM 4c or 4e, and the induction of apoptosis was assessed by labeling cell membrane phosphatidylserine with annexin V-FITC and cell DNA with PI combined with flow cytometry. The results of the apoptosis evaluation were in good agreement with the results from the MTT test. Both compounds induced apoptosis more efficiently in cancer HepG2 cells than in normal WI-38 cells: 56% and 20% of cells remained alive after incubation with 4c or 4e, respectively ( Figure 3C,D). The tested compound 4c did not induce apoptosis in normal fibroblasts at a concentration of 25 µM, whereas cells were more sensitive to 4e ( Figure 3E,F). Thus, 4e demonstrated cytotoxicity against both normal and cancer cells. Taken together, the results demonstrate that the bispidine derivatives 4c and 4e possess strong cytotoxic activity and can induce apoptosis in cancer cells, while normal cells are less sensitive to their action.

Polyamines Enhance the Toxicity of Diazabicyclononane Derivative 4e toward Cancer Cells
The study of cell viability and apoptosis induction demonstrated that compound 4e could induce a higher rate of cancer cell death than 4c, although no correlation was revealed between the ability to enhance PA catabolism ( Figure 2) and cell viability or apoptosis induction ( Figure 3) of the tested compounds. Therefore, subsequent experiments were performed with the most active compounds, 4c and 4e.
The degradation of PAs is commonly reduced and abnormal in cancer cells [20,21], while the products of PA oxidation are known to be able to induce cancer cell death by apoptosis [40,41]. To determine whether the cytotoxic activity of diazabicyclononane derivatives is truly associated with the activation of PA catabolism, we incubated HepG2 or WI-38 cells with a maximum nontoxic concentration (MNTC) of 2 µM for compounds 4c and 4e in the presence of 1 or 10 µM Spm or Spd. The controls were incubated with 4c and 4e and with each of PAs separately.
Both PAs in a single compound treatment promoted the viability of HepG2 and WI-38 cells in the MTT test ( Figure 4A,B,E,F). The combination of nontoxic concentrations of 4e and 4c with PAs resulted in different effects on cell viability. Namely, the 4e + 1 µM Spd/Spm combination caused 50% cell death in HepG2 cells ( Figure 4A,E). The combination of 4e + 10 µM Spd/Spm demonstrated even greater cytotoxicity, leaving only 10-20% of living cells. This effect suggests a PA-dose-dependent enhancement of the cytotoxicity of 4e. In contrast, Spd/Spm-induced WI-38 proliferation was slightly lowered by coincubation with 4e ( Figure 4B,F). Compound 4c was not able to induce any significant reduction in cell viability in the presence of PAs ( Figure 4C,D,G,H). Moreover, 4c did not affect the viability of either cancer or normal cells in PA-containing media. These results suggest that the cytotoxic activity of 4c, which is shown in Figure 3, is not associated with the catabolism of PAs, and the addition of PAs to 4c-treated cells did affect cell viability.  To reveal the type of cell death induced by the combination of 4e and PAs, cells were labeled with Annexin-V and FITC and subjected to flow cytometry. The results were in good agreement with the MTT test. Significant induction of HepG2 cell apoptosis was observed after incubation with 4e in the presence of each PA ( Figure 5A,C). Higher concentrations of Spd or Spm could induce a greater degree of cell death through apoptosis. However, incubation of normal WI-38 cells with PAs did not have any significant influence on cell death ( Figure 5B,D). The results of this experiment demonstrate that diazabicyclononane derivative 4e is able to induce apoptosis at the highest nontoxic concentration in the presence of PAs in HepG2 cancer cells but not in normal WI-38 fibroblasts. of Spd or Spm could induce a greater degree of cell death through apoptosis. However, incubation of normal WI-38 cells with PAs did not have any significant influence on cell death ( Figure 5B,D). The results of this experiment demonstrate that diazabicyclononane derivative 4e is able to induce apoptosis at the highest nontoxic concentration in the presence of PAs in HepG2 cancer cells but not in normal WI-38 fibroblasts. To reveal the type of cell death induced by the combination of 4e and PAs, cells were labeled with Annexin-V and FITC and subjected to flow cytometry. The results were in good agreement with the MTT test. Significant induction of HepG2 cell apoptosis was observed after incubation with 4e in the presence of each PA ( Figure 5A,C). Higher concentrations of Spd or Spm could induce a greater degree of cell death through apoptosis. However, incubation of normal WI-38 cells with PAs did not have any significant influence on cell death ( Figure 5B,D). The results of this experiment demonstrate that diazabicyclononane derivative 4e is able to induce apoptosis at the highest nontoxic concentration in the presence of PAs in HepG2 cancer cells but not in normal WI-38 fibroblasts.

Discussion
Our study revealed biological activity among novel bispidine derivatives. The tested compounds can activate PA catabolism, and two of them, 4e and 4c, demonstrate pronounced toxic activity toward cancer cells. Our experiment also revealed a PA-dependent enhancement of the cytotoxicity of the most active bispidine derivative 4e. The results of our study demonstrate that tumor selectivity and potent anticancer activity for this compound could be achieved in combination with PAs.
Our study is in accordance with the recent work by Kanamori et al. [42], who demonstrated that exogenous bovine serum amine oxidase and PA could induce apoptosis in cancer but not normal cells. However, we aimed to enhance the same PA-degrading pathway using small molecule drug substances to activate endogenous enzymes of PAs catabolism.
Previously, it was shown that exogenous PAs can promote tumor and normal cell proliferation [43] and provide a cytoprotective action for normal fibroblasts [44]. In accordance, we demonstrated the promotion of WI-38 and HepG2 viability after cell incubation with PAs ( Figure 4A,B). The viability of normal WI-38 cells in the presence of combinations 4e + Spd/Spm and 4c + Spd/Spm was not significantly different from that induced by PAs alone. These results suggest that the tested bispidine derivatives had no effect on the viability of normal cells. In case of cancer HepG2 cells, these two compounds demonstrated opposite effects: 4c, which was not active at MNTC, caused no cell death when added in combination with PAs, while 4e cytotoxic activity was potentiated by exogenous PAs. The products of PA catabolism (mainly reactive oxygen species (ROS), such as hydrogen peroxide and acrolein) are capable of inducing cell death via apoptosis [25,26,32,33]. Apoptosis was the type of cell death in HepG2 cells when 4e was coincubated with Spd/Spm. The absence of cell death at the MNTC of the compound suggests that the initial cellular content of PAs in tumor HepG2 cells (Table 2) is not enough high to produce sufficient amounts of toxic metabolites of PA decomposition to cause apoptosis. We believe that the addition of exogenous PAs resulted in their oxidation by bispidine-activated PA catabolic enzymes and the generation of cytotoxic products that could trigger cell death. Such an effect was specific only for 4e but not for 4c, despite its more pronounced ability to accelerate PA catabolism shown in regenerating rat liver homogenates.
We believe that the low transmembrane penetration or intracellular metabolism of 4c may be the reason for its low activity in living cells. The tested compounds were converted into complexes with β-cyclodextrins that are widely used in medicinal chemistry to improve water solubility and the transportation of hydrophobic drugs to the target cell membrane barriers. The penetration of such complexes inside the cell may occur via two main pathways [45]. The first is endo-and/or pinocytosis, which is typical for similar complexes regardless of the chemical composition. The second mechanism relies on the dissociation of complexes with the release of active substances that can pass through the cell membrane. This pathway depends on the physical and chemical properties of the active compound. The mechanism of cellular uptake of bispidines has not yet been reported. Compound 4e has some structural features that distinguish it from other compounds and may potentiate its activity. In particular, it contains nonaromatic piperazinyl-ethyl-side radicals at position N7, while all other tested substances include aromatic radicals at that position. This saturated diazacyclic fragment has a secondary amino group (Figure 6), is protonated and positively charged under physiological conditions due to the basic properties of piperazine, which are much stronger than those in imidazoles or pyridines [46]. Moreover, the positive charge is not delocalized in aliphatic piperazine ring unlike aromatic imidazole or pyridine rings. The saturation of the C-C bonds and positively charged nitrogen make 4e structurally more similar to natural aliphatic PAs ( Figure 6). This structural similarity may promote the interaction of 4e with natural PA transporters [47] and their penetration through the membrane after the dissociation of the bispidine-β-cyclodextrin complex. Thus, we can assume that the piperazine moiety unique to the lead compound 4e may facilitate its uptake by the cells, thus enhancing its intracellular toxic effect. imidazole or pyridine rings. The saturation of the C-C bonds and positively charged nitrogen make 4e structurally more similar to natural aliphatic PAs ( Figure 6). This structural similarity may promote the interaction of 4e with natural PA transporters [47] and their penetration through the membrane after the dissociation of the bispidine-β-cyclodextrin complex. Thus, we can assume that the piperazine moiety unique to the lead compound 4e may facilitate its uptake by the cells, thus enhancing its intracellular toxic effect. Another interesting observation is the selectivity of 4e toward the cancer cell line HepG2, which is obvious and needs to be explained. The concentrations of PAs in HepG2 and WI-38 cells are comparable (Table 2), according to the literature data, and are much lower than the amount of exogenously added PA. Thus, they cannot be responsible for the different effects of 4e on cell viability in these two cell lines. We suppose that the differences in PA metabolism enzyme activities in cancer and normal cells are more likely to  Another interesting observation is the selectivity of 4e toward the cancer cell line HepG2, which is obvious and needs to be explained. The concentrations of PAs in HepG2 and WI-38 cells are comparable ( Table 2), according to the literature data, and are much lower than the amount of exogenously added PA in our experiments. Thus, they cannot be responsible for the different effects of 4e on cell viability in these two cell lines. We suppose that the differences in PA metabolism enzyme activities in cancer and normal cells are more likely to be the reason for the observed selective cytotoxicity of 4e. The available data on PA metabolism characteristics in the experimental cell lines are summarized in Table 2. Additionally, normal hepatocytes and mesenchymal fibroblasts have been added for a better illustration, as not all the parameters for the experimental cell lines have been found in the literature. The absolutely zero transcription level of one PA catabolic enzyme, acetl-polyamine oxidase (APAO), in HepG2 cells was reported in earlier studies. This feature is one of the key differences between normal and tumor tissue metabolism [2,3,[20][21][22][23][24][25]. In contrast, spermine oxidase (SMOX) is more active in cancer HepG2 cells than in normal cells. The overexpression of this PA catabolic enzyme is often associated with cellular malignization [37] due to a double-edged role of ROS in tumor progression. Our compounds are potent activators of the PA catabolic pathway, and a rapid ROS storm may cause apoptosis rather than malignization and tumor promotion. In addition to direct activation, induction of the corresponding gene expression might also take place, although this hypothesis needs to be further verified.
The benzoylated and nonbenzoylated pairs did not demonstrate any pronounced drug/prodrug activity. Instead, adding benzoyloxim to the basic structures provided mostly greater activity toward PA catabolism and antiproliferative action.
As demonstrated in Figure 4, adding each of PAs as a substrate to boost toxicity provided convincing evidence for the potential of bispidines. Minimally toxic bispidines strongly synergized the potency of the two PAs in the tumor cell line but not in fibroblasts. Taken together, the results of the study demonstrate that synthesized bispidine derivatives can induce cancer cell death and the activation of PA catabolism. In the present study, we showed for the first time that the bispidine derivative 4e 3-(3-methoxypropyl)-7-[3-(1H-piperazin-1-yl)ethyl]-3,7-diazabicyclo[3.3.1]nonane can be subjected to PA-based drug development.

Reagents and Equipment
Primary amines, paraform and 1-Boc-piperidin-4-one (1a) were purchased from Aldrich (Louis Street, MO, USA). IR spectra were recorded on a Nicolet 5700 instrument between KBr plates. 1H and 13C NMR spectra were recorded on a JNM-ECA Jeol 400 spectrometer (frequencies 399.78 and 100.53 MHz, respectively) using DMSO-d6 solvent. The elemental analysis data were consistent with the calculated values. Column chromatography and thin-layer chromatography were carried out on alumina (Al 2 O 3 ) of the third degree of activity, and Rf compounds were given for this type of plate. The spots were developed in iodine vapors. The IR and NMR spectra for the synthesized compounds are presented in the Supplementary File.

Syntheses of Bispidine-9-Ones (Compounds 2a-2e)
1-(3-Hydroxypropyl)piperidin-4-one (1b). A total of 385 mL of absolute toluene and 16.26 g (0.707 mol) of metallic Na were placed in a three-necked flask equipped with a mechanical stirrer, reflux condenser with a calcium chloride tube, and dropping funnel. The reaction mixture was heated to 110 • C in an oil bath until the sodium dissolved (to obtain suspensions of sodium in toluene) for 25-30 min. Then, the reaction mixture was cooled to 75-80 • C, then 128 mL of methanol was added dropwise and it was stirred at 75 • C for 2 h. Afterwards, a straight condenser was attached to the flask, and the reaction mixture was heated. A solution of 174.61 g (0.707 mol) of diester was added using a dropping funnel at 110 • C, while the azeotropic mixture of toluene and methanol was simultaneously distilled off by direct distillation. The rate of dropping of the mixture of diester and methyl alcohol should be equal to the rate of distillation of the solvents. When the temperature reached 110 • C, becoming equal to the boiling point of toluene, the heating was stopped. While cooling (ice was used for cooling) and stirring, a solution of 301 mL of concentrated hydrochloric acid and 301 mL of distilled water was gradually added. The formed organic and aqueous layers were separated. The lower aqueous acidic layer was boiled at 100 • C for 8 h. The progress of the reaction was monitored using a solution of 1% FeCl 3 . The color changed to maroon. The solution was alkalized with NaOH to pH 12, extracted with chloroform, and dried over anhydrous MgSO 4 . The solvent was evaporated, and the residue was purified by column chromatography on Al 2 O 3 , benzene:dioxane = 5:1. A total of 21.04 g (19%) of 1-(3-hydroxypropyl)piperidin-4-one (1b) was obtained in the form of light yellow oil with R f = 0.65 (Al 2 O 3 , benzene:isopropanol = 7:1).

The Preparation of Regenerating Liver Homogenates
The regenerating rat liver was used as the source of the enzymes of PA catabolism [33]. The study was approved by the local Ethics Committee of RUDN University (protocol no. 17/09-2015). Six-month-old female Sprague-Dawley rats (RUDN vivarium) were subjected to partial hepatectomy according to standard surgical techniques [50]. Eighteen hours after the operation, regenerating livers were homogenized with an Omni MultiMix200 homogenizer (Omni, Kennesaw, GA, USA) in two volumes of ice-cold 50 mM Tris-HCl buffer, pH = 9.0, containing 0.05% SDS (Sigma-Aldrich, St. Louis, MO, USA) and protease inhibitors Complete™ Protease Inhibitor Cocktail (Roche, Basel, Switzerland). The homogenates were centrifuged at 10,000× g for 20 min at +4 • C, and the supernatant was used as a source of polyamine oxidases in their natural environment for the following in vitro evaluation of the action of bispidines on their activities.

Determination of Amine Oxidase Activity in Rat Liver Homogenates
Determination of amine oxidase activity in the presence of the tested compounds was performed in rat liver homogenates, as described previously [33]. A 30% liver homogenate in 50 mM Tris-HCl buffer, pH = 9.0, was incubated in 96-well plates for 1 h at 37

Cell Viability Testing
To determine the influence of the compounds on cell viability, an MTT assay was performed [52]. Briefly, cells were removed from culture flasks by trypsinization and seeded in 96-well plates (TPP, Trasadingen, Switzerland) at a concentration of 1 × 10 4 cells per well. In 24 h, synthesized compounds within the range of concentrations 1-50 µM were added and incubated for 72 h. Tetrazolium salt (3-(4,5-dimethyl-thiazol-2-yl)-2,5diphenyltetrazolium bromide, Serva, Heidelberg, Germany) solution (10 µL) was added to each well to reach a final concentration of 0.45 mg/mL and incubated at 37 • C for 4 hr. After incubation, the formazan crystals were dissolved in 100 µL dimethyl sulfoxide (DMSO), and the absorbance was measured at 570 nm with a CLARIOstar Plus multiplate reader (BMG Labtech, Ortenberg, Germany). The percentage of cell viability was calculated relative to nontreated control cells.
The maximum nontoxic concentration (MNTC) was considered the highest concentration of the compound that produced no statistically significant increase in cell viability. Compound concentrations that produce 50% of cell viability (IC50) were calculated from curves constructed by plotting cell viability (%) versus drug concentration (µM) [53]. To test the elevation of toxicity in the presence of PAs, cells were incubated with MNTC of 4c or 4e in the presence of 1 or 10 µM Spm·4HCI or Spd·3HCI (both from Sigma-Aldrich, St. Louis, MО, USA).

Detection of Apoptosis
To evaluate apoptosis, the incubated cells were re-suspended in phosphate saline buffer (Paneco, Moscow, Russia) and incubated with Annexin V-FITC and propidium iodide (PI) from a FITC Annexin V/Dead Cell Apoptosis kit (Life Technologies, Carlsbad, CA, USA), according to the manufacturer's protocol. The counting of 5 × 10 4 cells at each point was performed by flow cytometry with a MACS Quant Analyzer 10 (Miltenyi Biotec, Bergisch Gladbach, Germany) as it was previously described [54].

Statistics
SPSS 25 software (IBM SPSS Statistics, Armonk, NY) was used for statistical analysis. The data on PA catabolism influence belonged to the Gaussian distribution model, and parametrical methods were applied. The results are presented as mean ± standard error of mean (SЕM). The data on cytotoxicity and apoptosis evaluation did not belong to the Gaussian distribution model; non-parametrical methods were applied. Comparisons between compounds were performed using Mann-Whitney U test with correction for multiple testing according to Bonferroni method. Differences described by p ≤ 0.05 were considered significant. The results are presented as mean ± standard error of mean (SЕM).

Conclusions
The enhancement of PA catabolism in cancer cells has become a promising approach for the development of antitumor therapy. Synthetized bispidine derivative 4e 3-(3methoxypropyl)-7-[3-(1H-piperazin-1-yl)ethyl]-3,7-diazabicyclo[3.3.1]nonane demonstrated the ability to activate PA catabolism in regenerating rat liver homogenates. This compound could significantly decrease the viability of cancer cells in vitro. Adding Spd or Spm as a substrate of oxidation to boost toxicity, provided convincing evidence for the anticancer potential of bispidines. Minimally toxic bispidines strongly synergized the potency of the two PAs in the tumor cell line but not in normal fibroblasts. We can conclude that the lead compound 4e can become a potential anticancer drug substance which mechanism relies on the induction of PA catabolism in tumor cells.

Informed Consent Statement: Not applicable.
Data Availability Statement: The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.