Total Synthesis of Pulmonarin B and Design of Brominated Phenylacetic Acid/Tacrine Hybrids: Marine Pharmacophore Inspired Discovery of New ChE and Aβ Aggregation Inhibitors

A marine natural product, pulmonarin B (1), and a series of related tacrine hybrid analogues were synthesized and evaluated as cholinesterase (ChE) inhibitors. The in vitro ChE assay results revealed that 1 showed moderate dual acetylcholinesterase (AChE)/ butyrylcholinesterase (BChE) inhibitory activity, while the hybrid 12j proved to be the most potent dual inhibitor among the designed derivatives, being almost as active as tacrine. Molecular modeling studies together with kinetic analysis suggested that 12j interacted with both the catalytic active site and peripheral anionic site of AChE. Compounds 1 and 12j could also inhibit self-induced and AChE-induced Aβ aggregation. In addition, the cell-based assay against the human hepatoma cell line (HepG2) revealed that 1 and 12j did not show significant hepatotoxicity compared with tacrine and donepezil. Taken together, the present study confirmed that compound 1 was a potential anti-Alzheimer’s disease (AD) hit, and 12j could be highlighted as a multifunctional lead compound for anti-AD drug development.


Introduction
Alzheimer's disease (AD) is a chronic neurodegenerative disorder that has become the third leading death cause behind cancer and cardiovascular diseases. AD worsens over the time with the loss of memory, decline in language skills, deficits in cognitive functions, and severe behavioral problems [1]. In 2015, about 1.9 million people died of AD [2], and in 2017, an estimated 46.8 million AD patients were diagnosed worldwide [3]. More seriously, the number is expected to triple by 2050 with the aging of the global population [4].
Although the cause of AD is still poorly understood nowadays, many pathogenic hypotheses have been proposed over the last three decades [5,6]. Among them, cholinergic hypothesis is undoubtedly the earliest approved, which believes that increasing the level of acetylcholine (ACh) to enhance cholinergic neurotransmission in the brain is an efficacious approach for AD treatment [7]. In the brain, ACh can be hydrolyzed by two types of cholinesterase (ChE): acetylcholinesterase (AChE) and butyrylcholinesterase (BChE) [8]. AChE is the key enzyme for the termination of neurotransmission in cholinergic pathways via its rapid hydrolysis of ACh, almost 10 3 -fold more active than BChE [9]. Therefore, AChE inhibition is an effective approach for the symptomatic treatment for AD. As a result, Marine natural products (MNPs) have been proven to be an extremely important source for developing novel drugs [14,15]. As a special group of MNPs, marine halogenated metabolites possess a wide range of biological properties, such as antibacterial, antiviral, antitumor, anti-inflammatory, and neurological activities [16,17]. During our continued project for developing new ChE inhibitors [18][19][20], pulmonarin B (1, Figure 1), a brominated phenylacetic acid derivative isolated from the ascidian Synoicum pulmonaria by Seveson et al. [21], came to our line of sight. This dibrominated compound was reported to be an AChE inhibitor (IC50 = 36 µ M), and was considered as a marine hit for further studies [21]. It is important to note that the discovery of marine-derived AChE inhibitors was very limited according to the latest statistics of MNPs from 1984 to 2018 [22]. Thus, we recently prepared 1 using a synthetic approach that was different from that previously reported, and retested its anti-ChE effect. The assay results revealed that 1 was a moderate dual AChE/BChE inhibitor. In addition, it showed inhibitory activity against self-induced and AChE-induced Aβ1-42 aggregation and non-hepatotoxicity against HepG2 cells. However, the terminal quaternary ammonium group in 1 probably makes the compound's polarity too high to pass the blood-brain barrier. To further increase the bioactive profiles of 1, molecule hybridization was applied as a powerful strategy to assemble bioactive compounds. Tacrine as a versatile pharmacophore for the design of potent ChE inhibitors showed excellent dual inhibition on AChE/BChE with IC50 values at the nanomolar level [23][24][25]. Thus, a series of brominated phenylacetic acid/tacrine hybrids were designed by the fusion of tacrine to the terminal quaternary amine of 1. In the present study, we describe the total synthesis of 1 and its brominated phenylacetic acid/tacrine hybrids, as well as evaluate their anti-AD potential, including AChE/BChE inhibition, molecular docking, anti-Aβ aggregation, and cytotoxicity. Marine natural products (MNPs) have been proven to be an extremely important source for developing novel drugs [14,15]. As a special group of MNPs, marine halogenated metabolites possess a wide range of biological properties, such as antibacterial, antiviral, antitumor, anti-inflammatory, and neurological activities [16,17]. During our continued project for developing new ChE inhibitors [18][19][20], pulmonarin B (1, Figure 1), a brominated phenylacetic acid derivative isolated from the ascidian Synoicum pulmonaria by Seveson et al. [21], came to our line of sight. This dibrominated compound was reported to be an AChE inhibitor (IC 50 = 36 µM), and was considered as a marine hit for further studies [21]. It is important to note that the discovery of marine-derived AChE inhibitors was very limited according to the latest statistics of MNPs from 1984 to 2018 [22]. Thus, we recently prepared 1 using a synthetic approach that was different from that previously reported, and retested its anti-ChE effect. The assay results revealed that 1 was a moderate dual AChE/BChE inhibitor. In addition, it showed inhibitory activity against self-induced and AChE-induced Aβ 1-42 aggregation and non-hepatotoxicity against HepG2 cells. However, the terminal quaternary ammonium group in 1 probably makes the compound's polarity too high to pass the blood-brain barrier. To further increase the bioactive profiles of 1, molecule hybridization was applied as a powerful strategy to assemble bioactive compounds. Tacrine as a versatile pharmacophore for the design of potent ChE inhibitors showed excellent dual inhibition on AChE/BChE with IC 50 values at the nanomolar level [23][24][25]. Thus, a series of brominated phenylacetic acid/tacrine hybrids were designed by the fusion of tacrine to the terminal quaternary amine of 1. In the present study, we describe the total synthesis of 1 and its brominated phenylacetic acid/tacrine hybrids, as well as evaluate their anti-AD potential, including AChE/BChE inhibition, molecular docking, anti-Aβ aggregation, and cytotoxicity.

Chemistry
Given the particularly low yield of the key intermediate 3,5-(dibromo-4-methoxyphenyl)acetic acid in the protocol reported by Seveson et al. and the high price of the starting material 5-(dimethylamino) amylamine [21], the total synthetic route of pulmonarin B (1) was redesigned and successfully achieved with the agents in hand. As illustrated in Scheme 1, methyl 2-(4-methoxyphenyl) acetate (3) was prepared from the esterification of starting material 2-(4-methoxyphenyl) acetic acid (2). Ester 3 was dibrominated with NBS/FeCl 3 in acetonitrile to give compound 4, which was then hydrolyzed to yield acid 5. The yield from 2 to 5 is 43%, which was twofold higher than that (21%) reported by Severson et al. Next, amide 6 was prepared by a coupling reaction of 5 with tert-butyl (5-aminopentyl)carbamate. The deprotection of the N-Boc group of 6 in trifluoroacetic acid directly gave amine 7. Then, compound 8 was generated by the reductive amination of 7. Finally, N-methylation of 8 with CH 3 I enabled the full synthesis of pulmonarin B (1). The total yield of 1 was increased to 3.5%, which was over twofold what was reported in the literature (1.6%) [21]. Also, the present synthetic route afforded more pulmonarin B analogues by modifying the terminal amino group of intermediate 7.
The 1 H, 13 C NMR and MS spectrascopic data of the synthetic 1 were in agreement with those of the natural 1 [21].

Chemistry
Given the particularly low yield of the key intermediate 3,5-(dibromo-4-methoxyphenyl)acetic acid in the protocol reported by Seveson et al. and the high price of the starting material 5-(dimethylamino) amylamine [21], the total synthetic route of pulmonarin B (1) was redesigned and successfully achieved with the agents in hand. As illustrated in Scheme 1, methyl 2-(4methoxyphenyl) acetate (3) was prepared from the esterification of starting material 2-(4methoxyphenyl) acetic acid (2). Ester 3 was dibrominated with NBS/FeCl3 in acetonitrile to give compound 4, which was then hydrolyzed to yield acid 5. The yield from 2 to 5 is 43%, which was twofold higher than that (21%) reported by Severson et al. Next, amide 6 was prepared by a coupling reaction of 5 with tert-butyl (5-aminopentyl)carbamate. The deprotection of the N-Boc group of 6 in trifluoroacetic acid directly gave amine 7. Then, compound 8 was generated by the reductive amination of 7. Finally, N-methylation of 8 with CH3I enabled the full synthesis of pulmonarin B (1). The total yield of 1 was increased to 3.5%, which was over twofold what was reported in the literature (1.6%) [21]. Also, the present synthetic route afforded more pulmonarin B analogues by modifying the terminal amino group of intermediate 7. The 1 H, 13 C NMR and MS spectrascopic data of the synthetic 1 were in agreement with those of the natural 1 [21]. The synthetic routes for two series of compounds 10a-10h and 12a-12l were shown in Scheme 2. Briefly, intermediate 4 was reacted with appropriate amines to give the corresponding hybrids 10a-10h (Table 1). Similarly, the coupling reaction of 9f with different phenylacetic acids 11a-11l afforded the corresponding 12a-12l ( The synthetic routes for two series of compounds 10a-10h and 12a-12l were shown in Scheme 2. Briefly, intermediate 4 was reacted with appropriate amines to give the corresponding hybrids 10a-10h (Table 1). Similarly, the coupling reaction of 9f with different phenylacetic acids 11a-11l afforded the corresponding 12a-12l (

Chemistry
Given the particularly low yield of the key intermediate 3,5-(dibromo-4-methoxyphenyl)acetic acid in the protocol reported by Seveson et al. and the high price of the starting material 5-(dimethylamino) amylamine [21], the total synthetic route of pulmonarin B (1) was redesigned and successfully achieved with the agents in hand. As illustrated in Scheme 1, methyl 2-(4methoxyphenyl) acetate (3) was prepared from the esterification of starting material 2-(4methoxyphenyl) acetic acid (2). Ester 3 was dibrominated with NBS/FeCl3 in acetonitrile to give compound 4, which was then hydrolyzed to yield acid 5. The yield from 2 to 5 is 43%, which was twofold higher than that (21%) reported by Severson et al. Next, amide 6 was prepared by a coupling reaction of 5 with tert-butyl (5-aminopentyl)carbamate. The deprotection of the N-Boc group of 6 in trifluoroacetic acid directly gave amine 7. Then, compound 8 was generated by the reductive amination of 7. Finally, N-methylation of 8 with CH3I enabled the full synthesis of pulmonarin B (1). The total yield of 1 was increased to 3.5%, which was over twofold what was reported in the literature (1.6%) [21]. Also, the present synthetic route afforded more pulmonarin B analogues by modifying the terminal amino group of intermediate 7. The 1 H, 13 C NMR and MS spectrascopic data of the synthetic 1 were in agreement with those of the natural 1 [21]. The synthetic routes for two series of compounds 10a-10h and 12a-12l were shown in Scheme 2. Briefly, intermediate 4 was reacted with appropriate amines to give the corresponding hybrids 10a-10h (Table 1). Similarly, the coupling reaction of 9f with different phenylacetic acids 11a-11l afforded the corresponding 12a-12l (Table 2).

In Vitro Inhibition of AChE and BChE, and Structure-Activity Relationship (SAR) Analysis
At first, the in vitro AChE and BChE inhibitory activities of 1 were tested. From the results in Table 1, compound 1 showed comparable AChE inhibitory activity (IC 50 = 37.02 ± 2.11 µM) to the reported value of 36 µM [21]. In addition, it was also found to be a BChE inhibitor with an IC 50 value of 30.70 ± 1.44 µM. However, the ChE inhibitory activity of 1 was much weaker than that of the positive control tacrine (AChE, IC 50 = 0.159 ± 0.007 µM; BChE, IC 50 = 0.046 ± 0.002 µM). Aiming to improve the activity of 1, brominated phenylacetic acid/tacrine hybrids 10a-10h were prepared based on molecule hybridization. Compared with 1, all of these hybrids showed more potent ChE inhibitory activity, and clearly this improvement could be attributed to the introduction of tacrine moiety. In both enzyme bioassays, compound 10f with 7-carbon aliphatic spacer in the linker showed the best activity (AChE, IC 50 = 0.314 ± 0.010 µM; BChE, IC 50 = 0.053 ± 0.007 µM). The preliminary SAR study indicated that the ChE inhibitory activities of these compounds increased with the elongation of the diamine linker (from 10a to 10f) and then decreased (from 10f to 10h), and the 7-carbon diamino linkage seemed to be the optimal distance between the brominated phenylacetyl and 1,2,3,4-tetrahydroacridine moieties.

In Vitro Inhibition of AChE and BChE, and Structure-Activity Relationship (SAR) Analysis
At first, the in vitro AChE and BChE inhibitory activities of 1 were tested. From the results in Table 1, compound 1 showed comparable AChE inhibitory activity (IC50 = 37.02 ± 2.11 µ M) to the reported value of 36 µ M [21]. In addition, it was also found to be a BChE inhibitor with an IC50 value of 30.70 ± 1.44 µ M. However, the ChE inhibitory activity of 1 was much weaker than that of the positive control tacrine (AChE, IC50 = 0.159 ± 0.007 µ M; BChE, IC50 = 0.046 ± 0.002 µ M). Aiming to improve the activity of 1, brominated phenylacetic acid/tacrine hybrids 10a-10h were prepared based on molecule hybridization. Compared with 1, all of these hybrids showed more potent ChE inhibitory activity, and clearly this improvement could be attributed to the introduction of tacrine moiety. In both enzyme bioassays, compound 10f with 7-carbon aliphatic spacer in the linker showed the best activity (AChE, IC50 = 0.314 ± 0.010 µ M; BChE, IC50 = 0.053 ± 0.007 µ M). The preliminary SAR study indicated that the ChE inhibitory activities of these compounds increased with the elongation of the diamine linker (from 10a to 10f) and then decreased (from 10f to 10h), and the 7-carbon diamino linkage seemed to be the optimal distance between the brominated phenylacetyl and 1,2,3,4tetrahydroacridine moieties. Next, the substituent effect of benzene was investigated, and compound 10f was selected as a model compound for further structural modification. Consequently, 12 analogues (12a-12l) were prepared, and their bioassay results were shown in Table 1. Among these analogues, compound 12j showed the best ChE inhibitory activity (AChE, IC50 = 0.182 ± 0.006 µ M; BChE, IC50 = 0.064 ± 0.006 µ M), which was almost equal to that of tacrine. From the bioassay results, it was clear that different substituents had obvious impact on the AChE and BChE inhibitory activities. For example, compound 12j with 5-Br and 2-F showed increased inhibitory activity toward AChE by fourfold

In Vitro Inhibition of AChE and BChE, and Structure-Activity Relationship (SAR) Analysis
At first, the in vitro AChE and BChE inhibitory activities of 1 were tested. From the results in Table 1, compound 1 showed comparable AChE inhibitory activity (IC50 = 37.02 ± 2.11 µ M) to the reported value of 36 µ M [21]. In addition, it was also found to be a BChE inhibitor with an IC50 value of 30.70 ± 1.44 µ M. However, the ChE inhibitory activity of 1 was much weaker than that of the positive control tacrine (AChE, IC50 = 0.159 ± 0.007 µ M; BChE, IC50 = 0.046 ± 0.002 µ M). Aiming to improve the activity of 1, brominated phenylacetic acid/tacrine hybrids 10a-10h were prepared based on molecule hybridization. Compared with 1, all of these hybrids showed more potent ChE inhibitory activity, and clearly this improvement could be attributed to the introduction of tacrine moiety. In both enzyme bioassays, compound 10f with 7-carbon aliphatic spacer in the linker showed the best activity (AChE, IC50 = 0.314 ± 0.010 µ M; BChE, IC50 = 0.053 ± 0.007 µ M). The preliminary SAR study indicated that the ChE inhibitory activities of these compounds increased with the elongation of the diamine linker (from 10a to 10f) and then decreased (from 10f to 10h), and the 7-carbon diamino linkage seemed to be the optimal distance between the brominated phenylacetyl and 1,2,3,4tetrahydroacridine moieties. Next, the substituent effect of benzene was investigated, and compound 10f was selected as a model compound for further structural modification. Consequently, 12 analogues (12a-12l) were prepared, and their bioassay results were shown in Table 1. Among these analogues, compound 12j showed the best ChE inhibitory activity (AChE, IC50 = 0.182 ± 0.006 µ M; BChE, IC50 = 0.064 ± 0.006 µ M), which was almost equal to that of tacrine. From the bioassay results, it was clear that different substituents had obvious impact on the AChE and BChE inhibitory activities. For example, compound 12j with 5-Br and 2-F showed increased inhibitory activity toward AChE by fourfold Next, the substituent effect of benzene was investigated, and compound 10f was selected as a model compound for further structural modification. Consequently, 12 analogues (12a-12l) were prepared, and their bioassay results were shown in Table 1. Among these analogues, compound 12j showed the best ChE inhibitory activity (AChE, IC 50 = 0.182 ± 0.006 µM; BChE, IC 50 = 0.064 ± 0.006 µM), which was almost equal to that of tacrine. From the bioassay results, it was clear that different substituents had obvious impact on the AChE and BChE inhibitory activities. For example, compound 12j with 5-Br and 2-F showed increased inhibitory activity toward AChE by fourfold compared with 12a, and toward BChE by fourfold over 12f. Compounds 12b and 12l without Br substitution showed low anti-ChE activity in comparison with 12j.

Kinetic Study of AChE and BChE Inhibition
Herein, compound 12j was selected as a representative prototype for kinetic assay to obtain information on the mode of action and binding site of this series of analogues. The mechanisms of AChE and BChE inhibitions were analyzed by recording substrate concentration−enzyme velocity curves in the presence of different concentrations of 12j.
Graphical analyses revealed both increasing slopes (decreasing V max ) and increasing intercepts (K max ) at rising concentrations of 12j (Figures 2A and 3A). The equilibrium constant for binding with the free enzyme (K i ) was obtained from the slope against inhibitor concentration [26]. The estimated K i value of 12j was 97.50 nM for AChE ( Figure 2B), and that for BChE was 9.91 nM ( Figure 3B). These patterns revealed compound 12j as a linear mixed-type ChE inhibitor, indicating that 12j may not only bind to the catalytic active site (CAS), but also interact with the peripheral anionic site (PAS) of both enzymes. compared with 12a, and toward BChE by fourfold over 12f. Compounds 12b and 12l without Br substitution showed low anti-ChE activity in comparison with 12j.

Kinetic Study of AChE and BChE Inhibition
Herein, compound 12j was selected as a representative prototype for kinetic assay to obtain information on the mode of action and binding site of this series of analogues. The mechanisms of AChE and BChE inhibitions were analyzed by recording substrate concentration−enzyme velocity curves in the presence of different concentrations of 12j.
Graphical analyses revealed both increasing slopes (decreasing Vmax) and increasing intercepts (Kmax) at rising concentrations of 12j (Figure 2A and Figure 3A). The equilibrium constant for binding with the free enzyme (Ki) was obtained from the slope against inhibitor concentration [26]. The estimated Ki value of 12j was 97.50 nM for AChE ( Figure 2B), and that for BChE was 9.91 nM ( Figure  3B). These patterns revealed compound 12j as a linear mixed-type ChE inhibitor, indicating that 12j may not only bind to the catalytic active site (CAS), but also interact with the peripheral anionic site (PAS) of both enzymes.

Docking Study
As compound 12j showed the best inhibition toward AChE, it was selected for docking simulation. First, the ligand alkylene-linked bis-tacrine that was derived from the original X-ray structure of AChE (PDB ID: 5EI5) was redocked to the protein to validate the molecular modeling methodology, as shown in Figure S1 (Supporting Information). The docking result could well reproduce the crystal structure, which indicated that the method was suitable for a docking study of compared with 12a, and toward BChE by fourfold over 12f. Compounds 12b and 12l without Br substitution showed low anti-ChE activity in comparison with 12j.

Kinetic Study of AChE and BChE Inhibition
Herein, compound 12j was selected as a representative prototype for kinetic assay to obtain information on the mode of action and binding site of this series of analogues. The mechanisms of AChE and BChE inhibitions were analyzed by recording substrate concentration−enzyme velocity curves in the presence of different concentrations of 12j.
Graphical analyses revealed both increasing slopes (decreasing Vmax) and increasing intercepts (Kmax) at rising concentrations of 12j (Figure 2A and Figure 3A). The equilibrium constant for binding with the free enzyme (Ki) was obtained from the slope against inhibitor concentration [26]. The estimated Ki value of 12j was 97.50 nM for AChE ( Figure 2B), and that for BChE was 9.91 nM ( Figure  3B). These patterns revealed compound 12j as a linear mixed-type ChE inhibitor, indicating that 12j may not only bind to the catalytic active site (CAS), but also interact with the peripheral anionic site (PAS) of both enzymes.

Docking Study
As compound 12j showed the best inhibition toward AChE, it was selected for docking simulation. First, the ligand alkylene-linked bis-tacrine that was derived from the original X-ray structure of AChE (PDB ID: 5EI5) was redocked to the protein to validate the molecular modeling methodology, as shown in Figure S1 (Supporting Information). The docking result could well reproduce the crystal structure, which indicated that the method was suitable for a docking study of

Docking Study
As compound 12j showed the best inhibition toward AChE, it was selected for docking simulation. First, the ligand alkylene-linked bis-tacrine that was derived from the original X-ray structure of AChE (PDB ID: 5EI5) was redocked to the protein to validate the molecular modeling methodology, as shown in Figure S1 (Supporting Information). The docking result could well reproduce the crystal structure, which indicated that the method was suitable for a docking study of AChE. Since the redocking of the ligand tacrine that originated from BChE (PDB ID: 4BDS) to its protein could not well reproduce the crystal structure, the binding mode of 12j with BChE was not discussed in this study. As shown in Figure 4, 12j was obviously able to bind to the active pocket of AChE. Here, AChE (PDB: 5EI5) [27] was used in the docking analysis because of its high resolution. The docking score for 12j/AChE was −17.50. The tacrine moiety of 12j could be spatially located in CAS ( Figure 4A) and showed stacking interactions with Trp84 and Phe330 ( Figure 4B). Furthermore, the protonated nitrogen atom in the tacrine ring underwent cation interaction with residues Phe330 and Trp84, and also formed an H-bond with the key residue His440. The brominated benzene moiety was oriented in the PAS by forming stacking interactions with Try70, Try121, and Trp279. The H atom of the amide group serves as an H-bond donor for interacting with Tyr121. In addition, hydrophobic interactions between 12j and key residues also contributed to the affinity of 12j to AChE.
AChE. Since the redocking of the ligand tacrine that originated from BChE (PDB ID: 4BDS) to its protein could not well reproduce the crystal structure, the binding mode of 12j with BChE was not discussed in this study. As shown in Figure 4, 12j was obviously able to bind to the active pocket of AChE. Here, AChE (PDB: 5EI5) [27] was used in the docking analysis because of its high resolution. The docking score for 12j/AChE was −17.50. The tacrine moiety of 12j could be spatially located in CAS ( Figure 4A) and showed stacking interactions with Trp84 and Phe330 ( Figure 4B). Furthermore, the protonated nitrogen atom in the tacrine ring underwent cation interaction with residues Phe330 and Trp84, and also formed an H-bond with the key residue His440. The brominated benzene moiety was oriented in the PAS by forming stacking interactions with Try70, Try121, and Trp279. The H atom of the amide group serves as an H-bond donor for interacting with Tyr121. In addition, hydrophobic interactions between 12j and key residues also contributed to the affinity of 12j to AChE.

Inhibition of Self-Induced and AChE-Induced Aβ Aggregation
Studies showed that ChE inhibitors not only increased the levels of ACh in the brain, they also reduced and prevented the formation of Aβ aggregation [28]. Compounds 1 and 12j were assessed for their ability to inhibit self-induced and AChE-induced Aβ1-42 aggregation using the thioflavin T (ThT) fluorescence method. Tacrine and donepezil were used as positive reference compounds, and galantamine was included as a negative control. The results are shown in Table 2. The inhibitions of compound 1 and 12j at 10 µ M against self-induced Aβ1-42 aggregation were 29.78 ± 1.45% and 32.37 ± 0.62%, respectively, which were higher than that of donepezil (17.95 ± 0.77%). In addition, compounds 1 and 12j also exhibited a significant inhibitory effect on AChE-induced Aβ aggregation, with inhibition ratios of 27.60 ± 1.96% and 47.73 ± 4.35%, respectively. As for galantamine, no significant inhibition was observed in self-induced and AChE-induced Aβ1-42 aggregation assay. Table 2. Inhibition of self-induced and AChE-induced Aβ aggregation by compounds 1 and 12j in comparison with tacrine, donepezil, and galantamine.

Inhibition of Self-Induced and AChE-Induced Aβ Aggregation
Studies showed that ChE inhibitors not only increased the levels of ACh in the brain, they also reduced and prevented the formation of Aβ aggregation [28]. Compounds 1 and 12j were assessed for their ability to inhibit self-induced and AChE-induced Aβ 1-42 aggregation using the thioflavin T (ThT) fluorescence method. Tacrine and donepezil were used as positive reference compounds, and galantamine was included as a negative control. The results are shown in Table 2. The inhibitions of compound 1 and 12j at 10 µM against self-induced Aβ 1-42 aggregation were 29.78 ± 1.45% and 32.37 ± 0.62%, respectively, which were higher than that of donepezil (17.95 ± 0.77%). In addition, compounds 1 and 12j also exhibited a significant inhibitory effect on AChE-induced Aβ aggregation, with inhibition ratios of 27.60 ± 1.96% and 47.73 ± 4.35%, respectively. As for galantamine, no significant inhibition was observed in self-induced and AChE-induced Aβ 1-42 aggregation assay. Table 2. Inhibition of self-induced and AChE-induced Aβ aggregation by compounds 1 and 12j in comparison with tacrine, donepezil, and galantamine.

In Vitro Cytotoxicity toward HepG2 Cells
The main reason for the withdrawal of tacrine from the market is its hepatotoxicity. Thus, to verify the hepatotoxicity of 1 and 12j, a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay [29,30] on the human hepatoma cell line (HepG2) was carried out. Tacrine and donepezil were used as control drugs, and the IC 50 values of all of the tested compounds toward HepG2 cells were  Table 3. The results indicated that compounds 1 and 12j did not show significant cytotoxicity against HepG2 cell (cell viability > 50% up to 80 µM) compared with tacrine and donepezil.

AChE/BChE Inhibitory Assay
The AChE and BChE inhibitory activities of compounds were determined by using a slightly modified Ellman's method [30,31]. Electric eel AChE, equine serum BChE, 5,5 -dithiobis(2-nitrobenzoic acid) (DTNB), phosphate buffer solution (PBS, pH 8.0), acetylthiocholine (ATC) iodide, and butyrylthiocholine (BUC) iodide were purchased from Sigma-Aldrich (Steinheim, Germany). Tacrine was used as positive control. Enzyme solutions were prepared at 2.0 U/mL in 2-mL aliquots. The assay medium consisted of 10 µL of enzyme, 40 µL of PBS, 20 µL of 0.01 M DTNB, and 10 µL of tested compound. Assayed solutions of tested compounds were pre-incubated with corresponding ChE for 5 min. The reaction was initiated by the addition of 20 µL of 0.01 M substrate (ATC or BUC). The activity was determined by measuring the increase in absorbance at 410 nm at 37 • C in 2-min intervals using Tecan Spark multimode microplate reader (Mannedorf, Switzerland). The percentage of inhibition (I) was calculated from the measured data as follows: I = (A c − A i )/A c × 100%, where A i and A c represent the change in the absorbance in the presence of an inhibitor and without an inhibitor, respectively.

Kinetic Assay
Kinetic studies of inhibition on AChE and BChE were performed by using Ellman's method as described above. The concentrations of used substrates were 0.07813 µM, 0.1563 µM, 0.3125 µM, and 0.625 µM. Linear regression was used for calculation of Lineweaver−Burk plots, and all of the calculations were performed using GraphPad Prism 5.0 software (GraphPad Software, La Jolla, CA, USA).

Molecular Docking
Molecular docking studies were performed using the Autodock 4.2 program (The Scripps Research Institute, La Jolla, CA, USA) [32,33]. The crystal structure of AChE (PDB ID: 5EI5) was obtained from the Protein Data Bank after eliminating the inhibitor and water molecules. The 3D structure of the ligand was built, and performed MMFF94 minimization by using ChemBio3D Ultra 12.0 (CambridgeSoft Corporation, Cambridge, MA, USA). Using Autodock Tools 1.5.6, the preparation of the receptor was made by the addition of hydrogen atoms and Gasteiger charges, and finally an assignment of atomic types as AD4 type, and then autotorsion was used to define the rotatable bonds in the ligand preparation. The resulting enzyme structure was used as an input for the Autogrid program. Autogrid performed pre-calculated atomic affinity grid maps for each atom type in the ligand, plus an electrostatics map and a separate de-solvation map presented in the substrate molecule. The dimensions of the box were set to 60 × 60 × 60 with grid spacing of 0.375 Å. Rigid ligand docking was performed for the compounds. Docking calculations were carried out using the Lamarckian genetic algorithm (LGA). The proposed docking complex image was created by Pymol 1.5 (DeLano Scientific LLC, San Carlos, California, USA).
3.6. Determination of the Inhibitory Potency on Self-Aβ 1-42 Aggregation In order to investigate the inhibition of compounds on Aβ 1-42 self-aggregation, a ThT-based fluorometric assay was performed [34]. 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) pre-treated Aβ 1-42 (GL Biochem Ltd., Shanghai, China) was dissolved in DMSO to make a 200 µM stock solution. The stock solution was centrifuged at the speed of 13,500 rpm for 10 min. The above supernatant was used for experiments. The tested compounds were dissolved in DMSO at concentrations of 0.8 mM.
A screening assay for the tested compounds to inhibit A aggregation was performed by measuring ThT fluorescence emission. Compounds (2 µL) and 2 µL of 200 µM Aβ 1-42 were added into 76 µL of phosphate-buffered saline (PBS at pH 7.4) in a 96-well plate. After incubation for 24 h at room temperature, 80 µL of 5 µM of ThT solution (in 50 mM of glycine-NaOH at pH 8.5) was added to the reaction solution. Fluorescence emission was measured at 490 nm with an excitation wavelength of 450 nm on a Tecan Spark multimode microplate reader. Identical spectra were recorded by performing the independent experiments thrice. The fluorescence intensities were compared, and the % inhibition was calculated by the following equation: 100 − [(F i − F b )/(F o − F b ) × 100], where F i , F o and F b are the fluorescence intensities obtained for Aβ aggregation in the presence of inhibitors Aβ 1-42 and ThT; in the presence of Aβ 1-42 and THT but no inhibitors; and the blanks containing ThT only.

Determination of the Inhibitory Potency on Aβ 1-42 Aggregation Induced by AChE
The co-incubation experiment of Aβ 1-42 with AChE was performed by ThT bioassay according to a reported protocol [35]. HFIP pre-treated Aβ 1-42 (GL Biochem Ltd., shanghai, China) was dissolved in DMSO to make a 200 µM stock solution. The stock solution was centrifuged at the speed of 13,500 rpm for 10 min. The above supernatant was used for experiments. The tested compounds were dissolved in DMSO at concentrations of 10 mM, and diluted to 640 µM by phosphate-buffered saline (PBS at pH 7.4). A screening assay for tested compounds to inhibit A aggregation was performed by measuring ThT fluorescence emission. Compounds (2 µL) and 2 µL of 200 µM Aβ 1-42 and 20 µL AchE (2 µ/mL, in PBS at pH 8.0) were added into 76 µL of PBS (pH 8.0) in a 96-well plate. After incubation for 24 h at room temperature, 100 µL of 5 µM ThT solution (in 50 mM of glycine-NaOH at pH 8.5) was added to the reaction solution. Fluorescence emission was measured at 490 nm with an excitation wavelength of 450 nm on a Tecan Spark multimode microplate reader. Identical spectra were recorded by performing the independent experiments thrice. The fluorescence intensities were compared, and the % inhibition was calculated by the following equation: 100 − [(F i − F b )/(F o − F b ) × 100], where F i , F o and F b are the fluorescence intensities obtained for Aβ aggregation in the presence of inhibitors Aβ 1-42 , AChE, and ThT; in the presence of Aβ 1-42 , AChE, and ThT but no inhibitors; and the blanks containing ThT only.

Cytotoxicity Bioassay
The cytotoxicity of selected compounds was evaluated on HepG2 cells. Cells were inoculated into 96-well plates. After incubation for 24 h, the cells were treated with different concentrations of tested compounds for 24 h, and then were incubated with 10 µL of 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) at 37 • C for 2 h. The formazan dye product was measured by the absorbance at 490 nm on a Tecan Spark multimode microplate reader.

Conclusions
In summary, pulmonarin B (1) and a series of brominated phenylacetic acid/tacrine hybrids 10a-10h and 12a-12l were synthesized and evaluated for their anti-AD potential. Compound 1 was found to be a moderate dual AChE/BChE inhibitor and also showed inhibition on self-induced and AChE-induced Aβ aggregation, which established 1 as an interesting marine hit for further anti-AD study. Among these hybrids, the best result was obtained for compound 12j with IC 50 of 0.182 ± 0.006 µM for AChE and 0.064 ± 0.006 µM for BChE. The kinetic and molecular docking studies well confirmed 12j as a mixed-type AChE inhibitor. Moreover, 12j displayed the highest inhibitory activity against self-induced and AChE-induced Aβ aggregation. In addition, compound 12j did not show obvious hepatotoxicity in comparison with tacrine and donepezil. In the present study, the biological evaluation together with computational analyses demonstrated these newly designed brominated phenylacetic acid/tacrine hybrids as attractive lead compounds toward the discovery of multifunctional anti-AD drugs.