Effects of Sponge-Derived Alkaloids on Activities of the Bacterial α-D-Galactosidase and Human Cancer Cell α-N-Acetylgalactosaminidase

During a search for glycosidase inhibitors among marine natural products, we applied an integrated in vitro and in silico approach to evaluate the potency of some aaptamines and makaluvamines isolated from marine sponges on the hydrolyzing activity of α-N-acetylgalactosaminidase (α-NaGalase) from human cancer cells and the recombinant α-D-galactosidase (α-PsGal) from a marine bacterium Pseudoalteromonas sp. KMM 701. These alkaloids showed no direct inhibitory effect on the cancer α-NaGalase; but isoaaptamine (2), 9-demethylaaptamine (3), damirone B (6), and makaluvamine H (7) reduced the expression of the enzyme in the human colorectal adenocarcinoma cell line DLD-1 at 5 μM. Isoaaptamine (2), 9-demethylaaptamine (3), makaluvamine G (6), and zyzzyanone A (7) are slow-binding irreversible inhibitors of the bacterial α-PsGal with the inactivation rate constants (kinact) 0.12 min−1, 0.092 min−1, 0.079 min−1, and 0.037 min−1, as well as equilibrium inhibition constants (Ki) 2.70 µM, 300 µM, 411 µM, and 105 µM, respectively. Docking analysis revealed that these alkaloids bind in a pocket close to the catalytic amino acid residues Asp451 and Asp516 and form complexes, due to π-π interactions with the Trp308 residue and hydrogen bonds with the Lys449 residue. None of the studied alkaloids formed complexes with the active site of the human α-NaGalase.


Introduction
Marine organisms, in particular marine sponges, are inexhaustible source of metabolites with various biological activities. A large group of secondary metabolites isolated from marine sponges are alkaloids [1]. Marine sponges of the genera Latrunculia, Batzella, Prianos, Zyzzya are a rich source of alkaloids bearing a pyrrolo [4,3,2-de]quinoline skeleton [2]. Pyrroloquinoline alkaloids have shown a variety of biological activities, including antifungal and antimicrobial [3], antioxidant [4], antimalarial activity [5]; inhibition of the HIF-1α/p300 interaction [6]. Some makaluvamines demonstrated in vitro and in vivo cytotoxicity against various types of cancer [7,8], which was attributed to the inhibition of topoisomerase II [9]. It is shown that numerous pyrroloiminoquinone alkaloids, including synthetic analogues, are active, their mechanisms of action, pharmacological properties, and safety have shown that they have great potential for creating new anticancer drugs [10].
The next interesting small group of biologically active marine alkaloids having a rare 1H-benzo[de]-1,6-naphthyridine skeleton are aaptamines [11]. Aaptamines have been isolated from marine sponges mainly of the genus Aaptos. All these compounds have either an N-methylated or a non-N-methylated 1,6-naphthyridine core fused with a functionalized benzenoid unit. Some of aaptamines have been reported to have α-adrenoceptor

Collection and Identification of Sponge Material
We used freeze-dried samples of sponges A. aaptos (O45-113) and Z. fuliginosa (O9-407), collected during scientific cruises of R/V Academik Oparin, which are kept in the collection of G.B. Elyakov Pacific Institute of Bioorganic Chemistry, Far Eastern Branch, Russian Academy of Sciences. The sponge A. aaptos was collected at Tho Chu Island (9 • 18.5 N 103 • 29.6 E, at a depth of 6 m), South China Sea, Vietnam, in May 2013 during the 45th scientific cruise of R/V Academik Oparin. The sponge Z. fuliginosa was collected at Mid Islet (16 • 14 S 149 • 59 E, at a depth of 10 m), Eastern Australia, in July 1989 during the 9th scientific cruise of R/V Academik Oparin. Sponge samples were freeze-dried and stored in a refrigerator until used. The sponges were taxonomically identified by Dr. Vladimir B. Krasokhin.

Cytotoxic Activity Assays
DLD-1 cells (8 × 10 3 /200 µL) were seeded in 96-well plates for 24 h at 37 • C in a 5% CO 2 incubator. The cells were treated with compounds 1-9 at concentrations ranging from 0 to 20 µM for an additional 24 h. Subsequently, cells were incubated with 15 µL MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2Htetrazolium) reagent for 3 h, and the absorbance in each well was measured at 490/630 nm using a microplate reader "Power Wave XS" (Bio Tek, Winooski, VT, USA). All the experiments were repeated three times, and the mean absorbance values were calculated. The results are expressed in absorbance by the compound's treatment compared to the non-treated cells (control 1) and the compounds in medium (control 2). All the experiments were made in triplicate.

Cell Treatment with Alkoloids
DLD-1cells (5.0 × 10 5 ) were seeded in 6 mm dishes for 24 h at 37 • C in a 5% CO 2 incubator. Then the cells were treated with compounds 1-9 at concentrations ranging from 0 to 20 µM for an additional 24 h. Then the cells were harvested using 0.02% EDTA, after this 3 mL of cold 15 mM Tris-HCl (pH 7.0) was added.

Preparation of Cell Lysate
Every 3-4 days cells were rinsed in phosphate buffered PBS, detached from the tissue culture flask by 1× trypsin/EDTA solution, harvested with RPMI-1640 medium and centrifuge at 500 rpm for 3 min. The culture media was discarded and cells pellet was resuspended in EDTA/Tris solution and frozen at −80 • C.

Isolation and Purification of α-N-Acetylgalactosaminidase from Cell Lysates
Frozen lysates of cancer cells in the 15 mM Tris buffer, pH 7.0 containing 0.02% EDTA, defrosted and sonicated 10 times during 1 min with a break of 1 min. To remove the Biomedicines 2021, 9, 510 5 of 17 cellular detritus, the cell homogenate was centrifuged at 4 • C, 10,000 rpm, 30 min. The supernatant proteins were precipitated with 35% ammonium sulfate, kept during four hours at 4 • C to form a pellet. The protein pellet was collected by centrifugation (4 • C, 10,000 rpm, 30 min) and dissolved in 0.05 M sodium citrate buffer, pH 5.0. The supernatant proteins were precipitated with 75% ammonium sulfate, kept overnight at 4 • C to form a pellet. The protein pellet was collected by centrifugation (4 • C, 10,000 rev/min, 30 min) and dissolved in 0.05 M sodium citrate buffer, pH 5.0. Both extracts were dialyzed against the same buffer, centrifuged to separate the insoluble precipitate. α-NaGalase activity was tested in both extracts and was found only in the protein fraction precipitated with 75% ammonium sulfate. This supernatant was used in further work as partly purified enzyme. The purification quality was controlled by 12% Laemmli-SDS-PAGE-electrophoresis [47].

Production and Purification of Recombinant α-D-galactosidase
The recombinant wild-type α-D-galactosidase (α-PsGal) was produced as described earlier [48]. The plasmid DNA pET-40b(+) containing insertion of the gene from the marine bacterium Pseudoalteromonas sp. KMM 701 encoding α-PsGal was transformed in the Escherichia coli strain Rosetta (DE3). Heterological expression was carried out at optimal conditions as described earlier [49]. Purification of the recombinant α-PaGal was performed, according to described procedures [48]. The purification quality was controlled by 12% Laemmli-SDS-PAGE-electrophoresis [47].

Enzyme Assay
The activity of α-PsGal and α-NaGalase were determined by increasing the amount of p-nithrophenol (pNP). To assay the α-NaGalase activity 0.05 mL of cell extract and 0.1 mL of substrate pNPNA-α-Gal (8. Absorbance was measured at 400 nm. Results were read with a computer program Gen5 and treated with ExCel computer program. The unit of standard activity (U) was defined as the amount of the enzyme catalyzing the formation of 1 µmol of p-nitrophenol (ε 400 = 18,300 M −1 cm −1 ) per 1 min under the conditions indicated. Specific activity (A) was calculated as the enzyme activity (U) per 1 mg of protein. All calculations were based on reactions with consumption of 10% of the chromogenic substrate. The protein concentrations were estimated by the Bradford method with BSA as a standard [50]. Residual activity (%) was calculated with formula: Residual activity (%) = v/v 0 ×100, where v is the initial reaction rate of enzymes under the action of an inhibitor or other factors, v 0 is the initial reaction rate of enzymes in the absence of the influence of any factors. Buffer solutions of α-PsGal (0.1 U/mL) and α-NaGalase (0.01 U/mL) were used in the further experiments.

The Inhibitory Potency of the Compounds 1-9 for the α-PsGal and α-NaGalase
To preliminary test the inhibitory potency of the compounds 25 µL the enzyme solutions of α-PsGal in 0.05 M sodium phosphate buffer (pH 7.0) or α-NaGalase in 0.05 M sodium citrate buffer (pH 5.0) were preincubated for 25 min at 20 • C with 5 µL of water soluble or 2 µL of ethanol soluble test compounds at various concentrations, respectively (from 10 −3 M to 10 −7 M in a probe), to allow enzyme and inhibitor interaction. In the control sample, 5 µL of water or 2 µL of EtOH were added to 25 µL of enzyme solution. The enzyme reaction was initiated by adding 120 µL of the substrate pNP-α-Gal (3.3 mM) for α-PsGal or pNPNA-α-Gal (8.8 mM) for α-NaGalase in the respective buffer solutions. After 10 min for α-PsGal or 60 min for α-NaGalase the reaction was stopped by adding 150 µL of 1 M Na 2 CO 3 . The amount of pNP was quantified by spectrophotometric detection at 400 nm. Results were evaluated as the percent of residual activity as described above.

The Irreversibility of α-PsGal Inhibition by Alkaloids
To determine the reversibility of the α-PsGal activity inhibition by alkaloids 2, 3, 8, and 9 15 µL of 0.625 mM aqueous solution of the compounds 2 and 3 or 6 µL of 2.5 mM ethanol solution of the compound 8 and 9 were added to 75 µL of the enzyme solution in 0.05 M sodium phosphate buffer (pH 7.0). After 30 min of the mixture incubation the aliquot of 30 µL was taken and 120 µL of 3.3 mM pNPG solution were added to initiate the reaction; the reaction was stopped by addition of 150 µL of 1M Na 2 CO 3 . The α-D-galactosidase activity was determined, as described above. The remaining reaction mixture was dialyzed against 1 L of 0.02 M sodium phosphate buffer (pH 7.0) for 60 h at 4 • C. The buffer was changed 3 times during the dialysis. The enzyme activity was determined, as described above. A sample of α-PsGal treated with H 2 O or EtOH in the absence of the inhibitor (75 µL of the enzyme solution and 15 µL of H 2 O or 6 µL of EtOH, respectively) was dialyzed and used as a control for determination of the initial activity. The experiment was carried out in two replicates. The residual activity was calculated as described above.

Kinetic Studies on 2, 3, 8 and 9 against the α-PsGal
The kinetic parameters K i and k inact were determined by classic methods [51]. The inactivation of α-PsGal was monitored by incubation of the enzyme with different concentrations of alkaloids in 0.02 mM sodium phosphate buffer, pH 7.0, at 20 • C in total volume of 30 µL (25 µL and 5 µL of α-PsGal and alkaloids, respectively), which were placed in cells of 96 cells plate. After preincubation of enzyme-inhibitor mixture for 0.5, 5, 10, 15, 20, 25, or 30 min 120 µL the fraction of remaining enzyme activity was measured as a function of time by addition to the inactivation mixture 120 µL of a solution of pNP-α-Gal (3.3 mM). The pseudo-first-order rate constant of inactivation (k obs ) was determined for each inactivator concentration as the slope of the v/v 0 dependence on the incubation time in semilogarithmic coordinates. The Excel software was used for these calculations. The second order rate constants for the inactivation process were determined by fitting the dependences of the k obs values on the concentration of the inactivators to the Hill or Michaelis-Menten equations. An analysis of the graphs and the choice of models for calculation of inhibition constants (K i , mM) and inactivation rate constants (k inact , min -1 ) were performed with the Origin computer software.

Theoretical Models of α-D-Galactosidase Complexes with Aaptamine and Makaluvamine Alkaloids
The 3D model of α-D-galactosidase (GenBank: ABF72189.2) was built using the Molecular Operating Environment version 2020.09 package (MOE, 2020.09; Chemical Computing Group ULC, 1010 Sherbrooke St. West, Suite #910, Montreal, QC, Canada, H3A 2R7, 2020), as described earlier [39,40]. Structures of alkaloids were obtained from the PubChem database and optimized using the MOE program. The evaluation of structural parameters, contact structure analysis, molecular docking, and visualization of the results were carried out with the Dock and Ligand interaction modules in the MOE 2020.09 program. The results were obtained using the equipment of Shared Resource Center Far Eastern Computing Resource of Institute of Automation and Control Processes Far Eastern Branch of the Russian Academy of Sciences (IACP FEB RAS, available online: https://cc.dvo.ru (accessed on 30 November 2020).

Cytotoxic Effect of Alkaloids on Human Colorectal Adenocarcinoma Cell Line DLD-1
We have previously shown that the DLD-1 adenocarcinoma cell line expresses α-NaGalase better than other cell lines, and this enzyme was well characterized [37]. Therefore, for research the effect of alkaloids on the enzyme expression, we used the cell line DLD-1. We investigated the effect of sponge-derived alkaloids on α-NaGalase activity in human cancer cells using two assays: one was a treatment of DLD cells with alkaloids (in vitro); the other was a direct treatment with alkaloids of isolated and purified α-NaGalase from lysates of various cancer cell lines.
As a first step, the effect of compounds 1-9 on the viability of human colorectal carcinoma DLD-1 cells was tested by MTS assay to determine their non-toxic concentration. It was shown that compounds 1, 4-9 are non-cytotoxic against DLD-1 cells at the dose up to 20 µM. Compound 2 was weakly cytotoxic: the decrease in cell viability was 20% at 20 µM. The concentration of compound 3 that caused a 50% reduction in cell viability (IC 50 ) of DLD-1 cells was 17 ± 0.2 µM. DLD-1 cells possessed 100% viability after treatment with 5 µM concentration of compounds.

Inhibitory Potency of Aaptamine and Makaluvamine Classes of Alkaloids on the α-NaGalase of Cancer Cells
All alkaloids 1-9 were tested on their influence on the hydrolyzing activity of the α-NaGalase isolated from cells of human carcinoma DLD-1, HT-29 and HCT-116, breast cancer MDA-MB-231, melanoma SK-MEL-5 and RPMI-7951, and mouse healthy epidermal cells JB6 Cl41. Our experiments showed that none of the tested compounds 1-9 had any direct inhibitory effects on the activity not only against human cancer cell α-NaGalase as well against α-NaGalase from mouse healthy epidermal cells. However, isoaaptamine (2), 9-demethylaaptamine (3), damirone B (6) and makaluvamine H (7) reduced the expression of the enzyme in the human colorectal adenocarcinoma cell line DLD-1 from 100% to 64, 57, 52, and 50%, respectively, at concentration 5 µM (Figure 1). direct inhibitory effects on the activity not only against human cancer cell α-NaGalase as well against α-NaGalase from mouse healthy epidermal cells. However, isoaaptamine (2), 9-demethylaaptamine (3), damirone B (6) and makaluvamine H (7) reduced the expression of the enzyme in the human colorectal adenocarcinoma cell line DLD-1 from 100% to 64, 57, 52, and 50%, respectively, at concentration 5 μM (Figure 1).  We observed a similar effect after treatmant with DLD-1 cells with sulfated polysaccharide fucoidan from brown alga Fucus evanescens [37]. Fucoidan significantly reduced the expression of the enzyme, but did not directly inhibit it. Currently, the information about correlation of activity or expression of α-NaGalase with activation of receptors or proteins of signaling pathways, responsible for carcinogenesis, is absent in the literature.

Inhibitory Potency of Aaptamine and Makaluvamine Classes of Alkaloids on the α-PsGal of the Marine Bacterium
According to the modern classification of carbohydrate active enzymes (CAZy), the α-PsGal belongs to the GH36 family and α-NaGalase from human cancer cells belongs to the GH27 family. These enzymes of the GH27 and GH36 families are related, they share a common gene and form clan D, the structure of their active center is very homologous [28]. Accordingly, we decided to test whether alkaloids will inhibit an marine bacterial enzyme closely related. We hoped that effective α-PsGal inhibitors would inhibit the highly malignant α-NaGalase enzyme in cells. All alkaloids 1-9 were tested on the hydrolyzing activity of the α-PsGal isolated from a marine bacterium Pseudoalteromonas sp. KMM 701. There was no notable activity against α-PsGal for aaptamine (1), aaptanone (4), damirones A (5) and B (6), and makaluvamine H (7) even at a high concentration of inhibitors (Table 1).  The tested alkaloids to varying degrees inhibited the GH36 α-PsGal. It is interesting to note that pentacyclic guanidine alkaloids from the sponge Monanchora pulchra inhibited only marine microbial enzymes (the α-PsGal and 1,3-exo-β-D-glucanase from marine fungus Chaetomium indicum), but did not act on eukaryotic 1,3-endo-β-D-glucanase from marine mollusk Spisula sacchalinensis and the GH109 family α-NaGalase from marine bacteria Arenibacter latericius KMM 426 T [40,52].
Of all tested alkaloids, only four compounds showed any inhibitory effects on the activity of the bacterial α-PsGal. They are isoaaptamine (2), 9-demethylaaptamine (3), makaluvamine G (8), and zyzzyanone A (9) ( Table 1). It was shown that inhibitors 2, 3, 8, and 9 irreversibly inactivate the α-PsGal. The irreversible inhibition was confirmed by the dialysis experiment. The activity of the enzyme did not recover after dialysis against the buffer solution for 60 h, but the enzyme in the absence of the inhibitor keeps 100% activity during this dialysis process. Studies on 2, 3, 8, and 9 against the α-PsGal

Kinetic
To investigate a possible mode of interaction of these alkaloids with the α-PsGal, kinetic studies were done. The results of kinetic studies of the α-PsGal inactivation by the action of the spongean alkaloids 2, 3, 8, and 9 are shown in Figure 2. Because of the α-PsGal inactivation developed relatively slowly, within a few minutes under experimental conditions, the inhibitory activity of the compounds can be described by the inactivation rate constant (k inact , min -1 ) and equilibrium inhibition constant K i [53]. The kinetic Equation (1) describes the irreversible slow inhibition of the α-PsGal under the action of alkaloids 2, 3, 8, and 9: where E is enzyme; I is inhibitor; n is Hill's coefficient of cooperativity. Studies on 2, 3, 8, and 9 against the α-PsGal

Kinetic
To investigate a possible mode of interaction of these alkaloids with the α-PsGal, kinetic studies were done. The results of kinetic studies of the α-PsGal inactivation by the action of the spongean alkaloids 2, 3, 8, and 9 are shown in Figure 2. Because of the α-PsGal inactivation developed relatively slowly, within a few minutes under experimental conditions, the inhibitory activity of the compounds can be described by the inactivation rate constant (kinact, min -1 ) and equilibrium inhibition constant Ki [53]. The kinetic Equation (1) describes the irreversible slow inhibition of the α-PsGal under the action of alkaloids 2, 3, 8, and 9: where E is enzyme; I is inhibitor; n is Hill's coefficient of cooperativity.
(a) (i) (a) (ii) Biomedicines 2021, 9, x FOR PEER REVIEW 10 of 18   Figure 2(i) (a, b, c, d). The inactivation rate constant (k obs ), as can be seen in Figure 2(ii) (a, b, c, d) is inhibitor concentration dependent and a gradual increase in inhibitor concentration increases the inactivation rate constant. The nonlinear S-shaped curves of k obs dependence on a concentration of compounds 2, 3, 8, and 9 (Figure 2(ii)) mean that inhibition of the α-PsGal by these slowly binding irreversible inhibitors is cooperative and occurs in two stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of k obs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true K i and k inact , as well as Hill's cooperativity coefficients are given in Table 2.
Of the group aaptamine alkaloids, only isoaaptamine (2) and 9-demethylaaptamine (3) showed inhibitory activity against the α-PsGal. As can be seen in Table 2 isoaaptamine (2) (K i = 2.7 µM) is significantly more active if compared with 9-demethylaaptamine (3) (K i = 300 µM), although their inactivation rate constants are almost equal (k inact 0.12 and 0.092 min −1 , respectively). More active inhibitor 2 shows positive cooperative binding with α-PsGal with n > 1 (n = 2.17), i.e., binding of one molecule of 2 with the enzyme increases enzyme affinity for other molecules of this inhibitor. In contrast, less active 3 displays negative cooperative binding with n < 1 (n = 0.84); binding of one molecule of 3 decreases enzyme affinity for other molecules of 3. Structurally, these inhibitors differ only slightly, compound 3 is the non-N1-methylated analogue of compound 2. A substitution pattern in a benzenoid part of compounds 2 and 3 is the same; both inhibitors have a hydroxyl group at C-9 position in contrast to compounds 1 and 4. The increased activity of 2 in comparison with 3 is presumably the consequence of the presence of the methyl group at N1 of a 1,6-naphthiridine core. Thus, the presence of the hydroxyl group at C-9 and the methyl group at the N-1 in a benzo-1,6-naphthiridine core are important factors for α-PsGal inhibition. Such observation was reported earlier for 2, which showed higher activity in inhibition of sortase A, compared with 1 and 3 [18]. Isoaaptamine (2) also was more active than aaptamine (1) in inhibiting β-1,3-glucanase of marine mollusks [20]. Small structural differences in aaptamine (1) and aaptanone (4) ( Table 2) in comparison with 2 and 3 resulted in loss of their inhibitory potency on the α-PsGal. aaptamine (1) stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of kobs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true Ki and kinact, as well as Hill's cooperativity coefficients are given in Table 2. isoaaptamine (2) stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of kobs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true Ki and kinact, as well as Hill's cooperativity coefficients are given in Table 2. stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of kobs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true Ki and kinact, as well as Hill's cooperativity coefficients are given in Table 2. stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of kobs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true Ki and kinact, as well as Hill's cooperativity coefficients are given in Table 2. stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of kobs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true Ki and kinact, as well as Hill's cooperativity coefficients are given in Table 2. ;damirone B (6) stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of kobs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true Ki and kinact, as well as Hill's cooperativity coefficients are given in Table 2. stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of kobs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true Ki and kinact, as well as Hill's cooperativity coefficients are given in Table 2. stages: (1) the formation reversible enzyme-inhibitor complex (E·I) and (2) irreversible inhibition of the enzyme in this complex (E-I). The experimental dependences of kobs on the concentration of the compounds 2, 3, 8, and 9 (Figure 2(ii)) more exactly are approximated by the Hill's Equation (2). Values of true Ki and kinact, as well as Hill's cooperativity coefficients are given in Table 2. Among the test alkaloids of the makaluvamine group, makaluvamine G (8) and zyzzyanone A (9) showed low inhibitory activity against the α-PsGal. Nonetheless, zyzzyanone A (9) (K i = 105 µM) is more active than makaluvamine G (8) (K i = 411 µM), but inactivation rate of 9 (0.037 min −1 ) is almost twice lower that of 8 (0.079 min −1 ) ( Table 2). The methylaminoethyl side chain and/or rigid planar dipyrroloquinone core of 9, apparently, slowly penetrate into the active site of α-PsGal reducing the rate of enzyme inactivation. More active inhibitor 9 shows higher coefficient of cooperativity (n = 4.1) than less active inhibitor 8 (n = 2.5). Comparing structures of alkaloids from the makaluvamine group it can be seen that both metabolites 8 and 9 have a hydroxyphenyl ring in their structure. Probably, the increased electron density of the hydroxyphenyl ring would facilitate non-covalent π-π interactions of these compounds with protein residues in the catalytic or allosteric sites of the enzyme. Earlier it has been shown that pentacyclic guanidine alkaloids from the sponge Monanchora pulchra also are slow-binding irreversible inhibitors of the α-PsGal without the formation of covalent bonds with amino acids of the enzyme active center [40].

Theoretical Models of α-D-Galactosidase Complexes with Aaptamine and Makaluvamine Alkaloids
To support the active-site-directed nature of the α-D-galactosidase inactivation and assess possible binding sites for inhibitors a molecular docking of the active alkaloids 2, 3, 8, and 9 was performed on the predicted 3D structure of the α-D-galactosidase active center [54]. Homology model of the α-PsGal 3D-structure was constructed earlier by molecular docking in the MOE program [54] using the crystal structure of the α-galactosidase from Lactobacillus acidophilus of GH36 family as template [55]. The active center of the α-Dgalactosidase is located in the central (β/α) 8 domain, Asp451 and Asp516 are the catalytic nucleophile and acid/base residues, respectively. Another highly conserved residue in the active center is Trp308. The enzyme α-PsGal is a typical O-glycoside hydrolase of the GH36 family. It was previously shown that its molecule consists of two identical subunits [29,30]. The computer model of the enzyme has previously been successfully used for the analysis of α-PsGal complexes with low molecular weight inhibitors [39,40,54].
Molecular docking of the alkaloids isoaaptamine (2), 9-demethylaaptamine (3), makaluvamine G (8), and zyzzyanone A (9) with α-PsGal showed that all studied compounds form complexes with the active site of the enzyme. The key interactions of these inhibitors with the active site of the α-D-galactosidase are shown in Figure 3. The complexes of the alkaloids with the α-PsGal active center are stabilized by π-π interactions with functionally important residue Trp308. Moreover, all alkaloids form H-bonds with residue Lys449. As can be seen from superposition of complexes of compound 2, 3, 8, 9, and D-Gal with the active center of the enzyme (Figure 4), the alkaloids shared space in the active center with D-Gal. This indicates the active-site-directed nature of the α-D-galactosidase inactivation by the tested compounds. Although the alkaloids 2, 3, 8, and 9 did not directly interact with catalytic residues Asp451 and Asp516, tricyclic parts of alkaloids located deeply in the active center block the access of the substrate into the pocket of the active center to the catalytic residues Asp451 and Asp516.
None of the studied alkaloids formed complexes with the active site of human α-NaGalase. If, in the case of α-PsGal, isoaaptamine binds to the active site of the enzyme, then, in the case of α-NaGalase, binding sites of the alkaloid are located outside the active site of the enzyme ( Figure 5). Therefore, bound isoaaptamine has not decreased the activity of cancer α-NaGalase.
As can be seen from superposition of complexes of compound 2, 3, 8, 9, and D-Gal with the active center of the enzyme (Figure 4), the alkaloids shared space in the active center with D-Gal. This indicates the active-site-directed nature of the α-D-galactosidase inactivation by the tested compounds. Although the alkaloids 2, 3, 8, and 9 did not directly interact with catalytic residues Asp451 and Asp516, tricyclic parts of alkaloids located deeply in the active center block the access of the substrate into the pocket of the active center to the catalytic residues Asp451 and Asp516.   None of the studied alkaloids formed complexes with the active site of human α-NaGalase. If, in the case of α-PsGal, isoaaptamine binds to the active site of the enzyme,   None of the studied alkaloids formed complexes with the active site of human α-NaGalase. If, in the case of α-PsGal, isoaaptamine binds to the active site of the enzyme,   of alkaloids 2, 3, 8, 9, and D-galactose in the active site of α-D-galactosidase. Ligands are shown as sticks. Isoaaptamine (2) is shown in pink; 9-demethylaaptamine (3) is shown in cyan; makaluvamine G (8) is shown in yellow; zyzzyanone A (9) is shown in green; D-galactose is shown in white. The molecular surface of the catalytic domain binding site is shown in red (negative), blue (positive), and gray (neutral) classes.
None of the studied alkaloids formed complexes with the active site of human α-NaGalase. If, in the case of α-PsGal, isoaaptamine binds to the active site of the enzyme, then, in the case of α-NaGalase, binding sites of the alkaloid are located outside the active site of the enzyme ( Figure 5). Therefore, bound isoaaptamine has not decreased the activity of cancer α-NaGalase.

Conclusions
We used two related α-glycosidases of Clan D: α-N-acetylgalactosaminidase (α-NaGalase) of the GH27 family from different human cancer cells and α-galactosidase (α-PsGal) of the GH36 family from the marine bacterium Pseudoalteromonas sp. KMM 701 to study the effect of the well-known marine sponge metabolites with a therapeutic potential and clarify the possible mechanism of their inhibitory action. The well-characterized α-PsGal has been proved to be a suitable model for characterizing the novel properties of the alkaloids. Isoaaptamine (2), 9-demethylaaptamine (3), makaluvamine G (8), and zyzzyanone A (9) have been shown to be irreversible slow-binding inhibitors of the α-PsGal of the GH36 family. The inhibitory ability of these alkaloids depends on the chemical structure of their molecules. The presence of the hydroxyl group at C-9 position and the N1-methyl group on a benzo-1,6-naphthyridine skeleton are important factors for α-PsGal inhibition by aaptamine alkaloids. A hydroxyphenyl ring in makaluvamine G and zyzzyanone A is important structural feature for their activity against the α-PsGal. Our findings have demonstrated the potential of marine natural products to be a prototype of the α-galactosidase inhibitors. However, these highly active marine compounds act selectively on carbohydrases; they have no direct inhibitory effect on the activity of the cancer α-NaGalase of the GH27 family. Therefore, the expected properties, similar to those of α-PsGal, in the related eukaryotic enzyme (α-NaGalase) were not confirmed. It is interesting to note that the same compounds isoaaptamine (2), 9-demethylaaptamine (3), and two other compounds damirone B (6) and makaluvamine H (7) reduced the expression of the enzyme in the cell line DLD-1 at non cytotoxic concentration 5 µM. However, further in vitro and in vivo studies are needed to study the mechanism of molecular suppression of α-NaGalase in cancer cells.