Amaryllidaceae Alkaloids of Belladine-Type from Narcissus pseudonarcissus cv. Carlton as New Selective Inhibitors of Butyrylcholinesterase

Thirteen known (1–12 and 16) and three previously undescribed Amaryllidaceae alkaloids of belladine structural type, named carltonine A-C (13–15), were isolated from bulbs of Narcissus pseudonarcissus cv. Carlton (Amaryllidaceae) by standard chromatographic methods. Compounds isolated in sufficient amounts, and not tested previously, were evaluated for their in vitro acetylcholinesterase (AChE; E.C. 3.1.1.7), butyrylcholinesterase (BuChE; E.C. 3.1.1.8) and prolyl oligopeptidase (POP; E.C. 3.4.21.26) inhibition activities. Significant human BuChE (hBUChE) inhibitory activity was demonstrated by newly described alkaloids carltonine A (13) and carltonine B (14) with IC50 values of 913 ± 20 nM and 31 ± 1 nM, respectively. Both compounds displayed a selective inhibition pattern for hBuChE with an outstanding selectivity profile over AChE inhibition, higher than 100. The in vitro data were further supported by in silico studies of the active alkaloids 13 and 14 in the active site of hBuChE.


Introduction
Alzheimer's disease (AD), a progressive neurodegenerative brain disorder featuring memory loss and cognitive impairments, has been named after German psychiatrist Alois Alzheimer. AD is the most common type of dementia among the elderly, generally diagnosed in individuals over the age of because of its relatively high concentration in the bulbs, the large bulb size, and their availability in large volumes. Galanthamine is reported as the major alkaloid of Narcissus pseudonarcissus, followed by haemanthamine. Haemanthamine also has interesting biological activities including inhibition of protein synthesis, antimalarial and antiretroviral activity, as well as cytotoxicity against various cancer cell lines [24][25][26]. Recently, some semi-synthetic derivatives of haemanthamine have been published as promising candidates for AD therapy [27].
As a part of our ongoing research on Amaryllidaceae alkaloids with implication to AD, this work reports the isolation of several such alkaloids from fresh bulbs of Narcissus pseudonarcissus cv. Carlton. The isolated alkaloids that had not been previously studied were submitted for biological evaluation to reveal their inhibition potential towards hAChE, hBuChE, and POP. In vitro data were further supported by investigating the compound's putative binding modes in the active site of hBuChE to display crucial interaction.
The new compound 13, named carltonine A, was obtained as a pale yellow oil. The HRMS of 13 showed a protonated molecular ion peak [ carbons. Therefore, the structure of 13 was established as depicted ( Figure 2). The assigned atoms are shown in Table 1. Carltonine B (14) was obtained as a pale yellow oil. The molecular formula was determined to be C26H28N2O3 from the protonated molecular ion peak [M + H] + found at m/z 417.2172 (417.2173 calcd for C26H29N2O3 + ) in the positive-ion HRMS. Due to the small quantity of sample, the 1 H NMR spectrum had poor resolution, but it looked similar to that of 13. In addition, the 13 C spectrum was in accordance with this observation (see Table 1). The only difference was the absence of signals assigned to two methoxy groups of the 1,2,4,5-tetrasubstituted aromatic fragment, which were replaced by signals corresponding to a strongly deshielded methylene group of a 1,3-dioxole moiety in the 1D spectra. The key correlations are presented in Figure 2.
Another new compound, carltonine C (15), was obtained as a yellowish amorphous solid. The 1D NMR spectrum, as well as the 2D NMR spectra of 15 and 13, were quite similar; however, the 13 C NMR data showed doubling of most carbon resonances and the absence of the N-methyl group of the central tertiary nitrogen. Moreover, the molecular formula was determined to be C44H49N3O5 from the protonated molecular ion peak [M + H] + found at m/z 700.3743 (700.3745 calcd for C44H50N3O5 + ). These prerequisites led to the identification of the structure shown in Figure 2. The framed substituents of the central tertiary nitrogen are identical; therefore, there is no reason for doubling the signals, as appeared in the 1D NMR data. Due to the steric hindrance of these two fragments, atropisomerism has been estimated as the reason for the doubling of some signals. The analysis was performed at ambient temperature and at 50 °C. The acquired results proved the suitability of this assumption. At ambient temperature, some signals were divided into two lines that broaden with temperature increase and eventually merge into a single line (see Supplementary Material). Unfortunately, the spectrum also revealed the signals of decomposition of this molecule. Carltonine B (14) was obtained as a pale yellow oil. The molecular formula was determined to be C 26 H 28 N 2 O 3 from the protonated molecular ion peak [M + H] + found at m/z 417.2172 (417.2173 calcd for C 26 H 29 N 2 O 3 + ) in the positive-ion HRMS. Due to the small quantity of sample, the 1 H NMR spectrum had poor resolution, but it looked similar to that of 13. In addition, the 13 C spectrum was in accordance with this observation (see Table 1). The only difference was the absence of signals assigned to two methoxy groups of the 1,2,4,5-tetrasubstituted aromatic fragment, which were replaced by signals corresponding to a strongly deshielded methylene group of a 1,3-dioxole moiety in the 1D spectra. The key correlations are presented in Figure 2.
Another new compound, carltonine C (15), was obtained as a yellowish amorphous solid. The 1D NMR spectrum, as well as the 2D NMR spectra of 15 and 13, were quite similar; however, the 13 C NMR data showed doubling of most carbon resonances and the absence of the N-methyl group of the central tertiary nitrogen. Moreover, the molecular formula was determined to be C 44  These prerequisites led to the identification of the structure shown in Figure 2. The framed substituents of the central tertiary nitrogen are identical; therefore, there is no reason for doubling the signals, as appeared in the 1D NMR data. Due to the steric hindrance of these two fragments, atropisomerism has been estimated as the reason for the doubling of some signals. The analysis was performed at ambient temperature and at 50 • C. The acquired results proved the suitability of this assumption. At ambient temperature, some signals were divided into two lines that broaden with temperature increase and eventually merge into a single line (see Supplementary Material). Unfortunately, the spectrum also revealed the signals of decomposition of this molecule.

Biological Activity Determination of Isolated Alkaloids
All the isolated compounds that had not been studied previously for their inhibition potential of cholinesterases, and which were obtained in sufficient amounts, were screened for their hAChE/hBuChE inhibition potency using a modified spectrophotometric method of Ellman et al. [36]. Furthermore, hAChE/hBuChE-active AAs were also studied for their ability to inhibit POP enzyme. Galanthamine and eserine were used as positive controls in the hAChE/hBuChE assays. Berberine was selected as a positive control when measuring POP inhibition. The results are summarized in Table 2. Moreover, the in vitro data are justified by docking studies proposing orientation of the top-ranked ligands within the hBuChE gorge.
The hAChE/hBuChE inhibitory activity of the AAs was initially screened at a concentration of 100 µM. Compounds displaying inhibition ability >50% against one or both cholinesterases at the screening concentration were selected for the determination of their IC 50 values (Table 2).
From the structural perspective, both AAs are endowed with the same core structure, differing only in the substitution at positions C-5 and C-6 , respectively ( Figure 2). The presence of a 1,3-dioxolane ring in 14 compared with its opened dimethoxybenzene analogue 13 is plausibly responsible for the almost 30 times drop in hBuChE inhibition activity. Both compounds showed a hBuChE selective inhibition pattern with outstanding selectivity index (SI) values higher than 100 (Table 2).  Our group has previously isolated similar compounds from fresh bulbs of Nerine bowdenii: 6-O-demethylbelladine (13) and 4 -O-demethylbelladine (14) (Figure 3) [37]. However, neither of these two alkaloids are substituted at position C-7 , and differ from each other by the absence of one methoxy group ( Figure 3) [37]. 4 -O-demethylbelladine (IC 50 = 30.7 ± 4.0 µM) displayed slightly better in vitro inhibition activity of hBuChE compared with galanthamine (IC 50 = 42 ± 1 µM). On the other hand, the compounds isolated within this study are more than 30 to 100 more potent hBuChE inhibitors, yielding a new structural lead scaffold that may be pursued in AD research. In the hAChE assay, all studied AAs displayed marginal inhibition potency ( Table 2). On the other hand, all newly isolated belladine-type alkaloids (13)(14)(15) demonstrated promising inhibition activity towards hBuChE (Table 2; Figure 3). Indeed, compounds 13 and 14 displayed IC50 values in the nanomolar range (hBuChE IC50 = 910 nM and 31 nM, for 13 and 14, respectively).
From the structural perspective, both AAs are endowed with the same core structure, differing only in the substitution at positions C-5´ and C-6´, respectively ( Figure 2). The presence of a 1,3dioxolane ring in 14 compared with its opened dimethoxybenzene analogue 13 is plausibly responsible for the almost 30 times drop in hBuChE inhibition activity. Both compounds showed a hBuChE selective inhibition pattern with outstanding selectivity index (SI) values higher than 100 ( Table 2).
Our group has previously isolated similar compounds from fresh bulbs of Nerine bowdenii: 6-Odemethylbelladine (13) and 4´-O-demethylbelladine (14) (Figure 3) [37]. However, neither of these two alkaloids are substituted at position C-7´, and differ from each other by the absence of one methoxy group (Figure 3) [37]. 4´-O-demethylbelladine (IC50 = 30.7 ± 4.0 µM) displayed slightly better in vitro inhibition activity of hBuChE compared with galanthamine (IC50 = 42 ± 1 µM). On the other hand, the compounds isolated within this study are more than 30 to 100 more potent hBuChE inhibitors, yielding a new structural lead scaffold that may be pursued in AD research.  Since some of the alkaloids were only isolated on a small scale, only two (1 and 13) were able to be tested for POP inhibition. Alkaloid 13 demonstrated POP inhibition in the same range as berberine ( Table 2) [38]. The biological profiles of the other AAs ( Figure 1) have already been determined within previous studies on Amaryllidaceae plants [21,[39][40][41].

Docking Study of Carltonine A (13) and Carltonine B (14)
To reveal fundamental interactions for 13 and 14 within the hBuChE active site (PDB ID: 4BDS) [42], molecular docking studies were carried out, enabling better insight into the structural requirements for inhibition. Since both 13 and 14 are tertiary amines, which are protonated at physiological pH, they behave as pseudo-enantiomers. In this study, we attempted to predict more bioactive conformers of 13 and 14 based on their energy score and their topology within hBChE. To gain a more in-depth outlook into the overall ligand-enzyme complex, we are also displaying several comparative structure overlay between (R)-13-(R)-14 ( Figure S4A Figure 5A,B), the N-methylindoline moiety occupies the vicinity of the catalytic triad with a T-shaped π-π interaction close to Phe329 (5.0 Å, distance measured from ring-to-ring center). Trp82 faces the dimethoxybenzene ring by displacing π-π interaction, whereas the phenolic appendage is left somewhat vacant. In the (S)-13 pseudo-enantiomer-hBuChE complex ( Figure 4C,D), the N-methylindoline moiety is engaged in parallel π-π stacking with Trp82 (3.7 Å). Contrary to the vacant occupancy of the phenolic appendage in the (R)-13 pseudo-enantiomer, in the case of the (S)-13 pseudo-enantiomer, Phe329 (4.9 Å) and Trp231 (5.5 Å) anchor this moiety. The dimethoxybenzene ring is stabilized by π-anion contact to Asp70.
Taken together from all the in silico observations above, it can be deduced that the higher inhibition potency of 14 over 13 might be attributed to the presence of the 2H-1,3-benzodioxole moiety for its "extended" aromatic properties that, especially in the (R)-14 pseudo-enantiomer, broaden the range of hydrophobic interactions between the ligand and enzyme (Table 3). All the amino acids exhibiting in the vicinity of the ligands up-to 6.0 Å are rendered. Hydrogen atoms of amino acids are omitted for clarity. Catalytic triad residues are shown in yellow, and amino acid residues in blue (A and C). In 2D diagrams (B and D), crucial amino acid residues are displayed in different colours depending on the nature of the interaction (e.g., purple for π-π, orange for anion-π, dark green for van der Waals contact, and light green for conventional hydrogen bond).
Taken together from all the in silico observations above, it can be deduced that the higher inhibition potency of 14 over 13 might be attributed to the presence of the 2H-1,3-benzodioxole moiety for its "extended" aromatic properties that, especially in the (R)-14 pseudo-enantiomer, broaden the range of hydrophobic interactions between the ligand and enzyme. Hydrogen atoms of amino acids are omitted for clarity. Catalytic triad residues are shown in yellow, and amino acid residues in blue (A,C). In 2D diagrams (B,D), crucial amino acid residues are displayed in different colours depending on the nature of the interaction (e.g., purple for π-π, orange for anion-π, dark green for van der Waals contact, and light green for conventional hydrogen bond).  In 2D diagrams (B and D), crucial amino acid residues are displayed in different colours depending on the nature of the interaction (e.g., purple for π-π, orange for anion-π, dark green for van der Waals contact, light green for conventional hydrogen bond).

General Experimental Procedures
All solvents were treated using standard techniques before use. All reagents and catalysts were purchased from Sigma Aldrich, Czech Republic and used without purification. The NMR spectra were obtained in CDCl3 and CD3OD at ambient temperature on a VNMR S500 (Varian, Palo Alto, CA, USA) spectrometer operating at 500 MHz for 1 H and 125.7 MHz for 13 C. Chemical shifts were recorded as δ values in parts per million (ppm) and were indirectly referenced to tetramethylsilane Hydrogen atoms of amino acids are omitted for clarity. Catalytic triad residues are shown in yellow and amino acid residues in blue (A,C). In 2D diagrams (B,D), crucial amino acid residues are displayed in different colours depending on the nature of the interaction (e.g., purple for π-π, orange for anion-π, dark green for van der Waals contact, light green for conventional hydrogen bond).

General Experimental Procedures
All solvents were treated using standard techniques before use. All reagents and catalysts were purchased from Sigma Aldrich, Czech Republic and used without purification. The NMR spectra were obtained in CDCl 3 and CD 3 OD at ambient temperature on a VNMR S500 (Varian, Palo Alto, CA, USA) spectrometer operating at 500 MHz for 1 H and 125.7 MHz for 13 C. Chemical shifts were recorded as δ values in parts per million (ppm) and were indirectly referenced to tetramethylsilane (TMS) via the solvent signal (CDCl 3 -7.26 ppm for 1 H and 77.0 ppm for 13 C; CD 3 OD-3.30 ppm for 1 H and 49.0 ppm for 13 C). Coupling constants (J) are given in Hz. For unambiguous assignment of 1 H and 13 C signals, 2D NMR experiments, namely gCOSY, gHSQC, gHMBC and NOESY, were measured using standard parameter settings and standard pulse programs provided by the producer of the spectrometer. ESI-HRMS were obtained with a Waters Synapt G2-Si hybrid mass analyzer of a quadrupole-time-of-flight (Q-TOF) type, coupled to a Waters Acquity I-Class UHPLC system. The EI-MS were obtained on an Agilent 7890A GC 5975 inert MSD operating in EI mode at 70 eV (Agilent Technologies, Santa Clara, CA, USA). A DB-5 column (30 m × 0.25 mm × 0.25 µm, Agilent Technologies, USA) was used with a temperature program: 100-180 • C at 15 • C/min, 1 min hold at 180 • C, and 180-300 • C at 5 • C/min and 5 min hold at 300 • C; and detection range m/z 40-600. The injector temperature was 280 • C. The flow-rate of carrier gas (helium) was 0.8 mL/min. A split ratio of 1:15 was used. TLC was carried out on Merck precoated silica gel 60 F254 plates. Compounds on the plate were observed under UV light (254 and 366 nm) and visualized by spraying with Dragendorff's reagent.

Plant Material
The fresh bulbs of Narcissus pseudonarcissus cv. Carlton were obtained from the herbal dealer Lukon Glads (Sadská, Czech Republic). Botanical identification was performed by Prof. L. Opletal. A voucher specimen is deposited in the Herbarium of the Faculty of Pharmacy in Hradec Králové under number: CUFPH-16130/AL-654.

Extraction and Isolation of Alkaloids
Fresh bulbs (30 kg) were minced and completely extracted with ethanol (EtOH) (96%, v/v, 3×) by boiling for 30 min under reflux; the combined extract was filtered and evaporated to dryness under reduced pressure. The crude extract (485 g) was acidified to pH 1-2 with 2% hydrochloric acid (HCl; 1 L) and the volume of the suspension was made up to 5 L with water. The suspension was filtered; the filtrate was defatted by diethyl ether (Et 2 O, 3 × 1.5 L), alkalized to pH 9-10 with a 10% solution of sodium carbonate (Na 2 CO 3 ) and extracted with ethyl acetate (EtOAc; 3 × 1.5L). The organic layer was evaporated to give 245 g of dark brown fluid residue. The alkaloid summary extract was again dissolved in 2% HCl (1000 mL), washed with Et 2 O (3 × 300 mL) and alkalized to pH 9-10 with 10% Na 2 CO 3 . The water layer was extracted with Et 2 O (4 × 350 mL) and chloroform (CHCl 3 ; 4 × 350 mL). Both Dragendorff positive parts were evaporated and pooled. A concentrated alkaloid extract (187 g) in the form of brown syrup was obtained.
Fraction III (16.3 g) was crystallized and recrystallized from EtOH and, finally, 6.39 g of galanthamine (6) was obtained.

hAChE and hBuChE Inhibition Assay
The hAChE and hBuChE activities were determined using a modified method of Ellman, as described [36,43,44]. Briefly, hAChE and hBuChE activities were determined using a modified method of Ellman, with acetylthiocholine iodide (ATChI) and butyrylthiocholine iodide (BuTChI) as substrates, respectively. Briefly, 8.3 µL of either blood cell lysate or plasma dilutions (at least six different concentrations), 283 µL of 5 mM 5,5 -dithiobis-2-nitrobenzoic acid (DTNB) and 8.3 µL of the sample dilution in dimethyl sulfoxide (DMSO) (40 mM, 10 mM, 4 mM, 1 mM, 0.4 mM, and 0 mM) were added to the semi-micro cuvette. The reaction was initiated by addition of 33.3 µL 10 mM substrate (ATChI or BuTChI). The final proportion of DTNB and substrate was 1:1. The increase of absorbance (∆A) at 436 nm for AChE and 412 nm for BuChE was measured for 1 min at 37 • C using a spectrophotometer (Synergy TM HT Multi-Detection Microplate Reader). Each measurement was repeated six times for every concentration of enzyme preparation. The % inhibition was calculated according to the following formula: where ∆A Bl is the increase of absorbance of the blank sample and ∆A Sa is the increase of absorbance of the measured sample. Inhibition potency of the tested compounds was expressed as an IC 50 value (concentration of inhibitor, which causes 50% cholinesterase inhibition).

POP Inhibition Assay
POP (EC 3.4.21.26) was dissolved in phosphate-buffered saline (PBS; 0.01 M Na/K phosphate buffer, pH 7.4, containing 137 mM NaCl and 2.7 mM KCl); the specific activity of the enzyme was 0.2 U/mL. The assay was performed in standard polystyrene 96-well microplates with a flat and clear bottom. Stock solutions of tested compounds were prepared in DMSO (10 mM). Dilutions (10 −3 to 10 −7 M) were prepared from the stock solution with deionized H 2 O; the control was performed with the same DMSO concentration. POP substrate, (Z)-Gly-Pro-p-nitroanilide, was dissolved in 50% 1,4-dioxane (5 mM). For each reaction, PBS (170 µL), tested compound (5 µL), and POP (5 µL) were incubated for 5 min at 37 • C. Then, substrate (20 µL) was added, and the microplate was incubated for 30 min at 37 • C. The formation of p-nitroanilide, directly proportional to the POP activity, was measured spectrophotometrically at 405 nm using a microplate ELISA reader (multimode microplate reader Synergy 2, BioTek Instruments, Winooski, VT, USA). The inhibition potency of tested compounds was calculated by nonlinear regression analysis and was expressed as an IC 50 value (concentration of inhibitor which causes 50% POP inhibition). All calculations were performed using GraphPad Prism software version 7.03 for Windows (GraphPad Software Inc., San Diego, California, USA) [44].

Molecular Modelling Studies
Two structures of hAChE and hBuChE were gained from RCSB Protein Data Bank-PDB ID: 4EY6 (crystal structure of hAChE) and 4BDS (crystal structure of hBuChE) [42,45]. All receptor structures were prepared by DockPrep function of UCSF Chimera (v. 1.4) and converted to pdbqt-files by AutodockTools (v. 1.5.6) [46,47]. Flexible residues selection was based on previous experience with either hAChE, hBuChE or the spherical region around the binding cavity [48][49][50]. Three-dimensional structures of ligands were built by Open Babel (v. 2.3.1), minimized by Avogadro (v 1.1.0) and converted to pdbqt-file format by AutodockTools [51]. The docking calculations were made by Autodock Vina (v. 1.1.2) with the exhaustiveness of 8 [52]. Calculation was repeated 20 times for each ligand and receptor and the best-scored result was selected for manual inspection.

Conclusions
The phytochemical study of the alkaloidal extract of Narcissus pseudonarcissus cv. Carlton resulted in the isolation of thirteen previously described AAs, and three new AAs of belladine-type, named carltonine A-C (13)(14)(15). Their structures were elucidated using a combination of NMR and MS analysis. Compounds isolated in sufficient quantity and not previously tested for their biological activities in relation to AD, were screened for their potential to inhibit hAChE, hBuChE and POP. Significant and selective hBuChE inhibitory activity was demonstrated by the newly described alkaloids carltonine A (13) and carltonine B (14) with IC 50 values of 0.91 ± 0.02 µM and 0.031 ± 0.001 µM, respectively. The in vitro results were justified by computational studies predicting plausible binding modes of compounds 13 and 14 in the active site of hBuChE. The new compounds exerted an interesting biological profile deserving further lead-optimization. The next step will be the development of an appropriate synthetic route leading to carltonine derivatives with follow-up preparation of semi-synthetic derivatives.  (15) in CDCl 3 at 50 • C; Figure S3-9: 13 C NMR spectrum of carltonine C (15) in CDCl 3 at 50 • C; Figure S4: Overlapped pseudo-enantiomers in the hBuChE active site and their topology difference.