Design, Synthesis, and Biological Evaluation of Notopterol Derivatives as Triple Inhibitors of AChE/BACE1/GSK3β for the Treatment of Alzheimer's Disease.

The pathogenesis of Alzheimer's disease (AD) is very complex, and there are many hypotheses. Therefore, the development of a multi-target-directed-ligand may be an effective therapeutic strategy. Our previous study showed that notopterol (a natural product from Notopterygium) is a dual BACE1/GSK3β inhibitor. In this study, we designed and synthesized 48 notopterol derivatives with furacoumarin as a scaffold in order to enhance their balanced AChE/BACE1/GSK3β inhibitory activity. Fortunately, 1c showed effective inhibitory activity against AChE (58.7% at 1.0 μM), BACE1 (48.3% at 20 μM), and GSK3β (40.3% at 10 μM). Furthermore, 1c showed good blood-brain barrier penetrability, suitable bioavailability, and oral safety. More importantly, 1c could ameliorate the impaired learning and memory in Aβ-induced AD mice. In conclusion, we reported the triple inhibitor of AChE/BACE1/GSK3β lead compounds based on a furocoumarin scaffold of notopterol for the first time, which provides a potential new strategy for the treatment of AD.

Introduction

Alzheimer’s disease (AD) is a common neurodegenerative disease, and the pathogenesis of AD is complex and involves multiple mechanisms including β amyloid (Aβ) deposition, tau hyperphosphorylation, neurofibrillary tangles, neuronal loss, and neurotransmitter dysfunction.[1,2] Although there are many new theories emerging, the current drug development of AD mainly focuses on Aβ cascade hypothesis, tau pathological hypothesis, and cholinergic hypothesis.[3] In recent years, many potential monoclonal antibodies or small molecule drugs have been developed for these three classical hypotheses. Unfortunately, most drugs have failed in clinical trials. Hatat et al. have synthesized a variety of compounds possessing both in vitro and in vivo activities toward three therapeutic targets (5-HT4R/5-HT6R, AChE) in the potential treatment of AD.[4] A lot of evidence supports the multi-target-directed-ligands approach as a tool to get around the problem of drug–drug interaction and to reduce the risk of toxicity that occurs during polypharmacotherapy.[2] Therefore, multi-target drugs may have more advantages in the prevention and treatment of AD.[5]

Acetylcholinesterase (AChE) is the classic target of the cholinergic hypothesis, and most of the AD drugs currently on the market are AChE inhibitors. There are currently only 4 acetylcholinesterase inhibitors (AChEIs) that have been marketed, as shown in Figure . Studies have found that long-term high-dose tacrine will produce gastrointestinal adverse reactions such as increased liver transaminase and nausea and vomiting. Lowering the dose can alleviate the occurrence of adverse reactions, but the effect of treatment will also significantly decrease, so it was withdrawn from the market.[6]

Figure 1

FDA approved AChEIs.

The main function of β secretase 1 (BACE1) is to cleave the Aβ precursor protein (APP) and further generate neurotoxic Aβ by γ secretase. It is a popular target for the design of drugs to inhibit Aβ production.[7] Therefore, people began to develop compounds against this target, such as LY2811376,[8] atabecestat,[9] and AZD3293 (lanabecestat).[10] These drugs have been found to reduce Aβ deposition in both short-term and long-term exposure in AD model animals and improve the cognitive deficits. However, these compounds have been found to cause increased hepatotoxicity or adverse reactions such as apathy and weight loss in clinical trials. The once-promising Verubecestat (MK-8931) developed by Merck was terminated in phase III clinical trials.

Although the level of Aβ in the brain and CSF of prodromal or mild to moderate AD patients exposed to Verubecestat decreased, it failed to improve their cognitive function.[11] BACE1 inhibitor E2609 (Elenbeestat), containing a guanidine scaffold,[12] was announced in September 2020. Due to poor benefits, the phase III clinical trial of E2609 was terminated. So far, all clinical trials of BACE1 inhibition have been wiped out (Figure ).

Figure 2

BACE1 inhibitors terminated in clinical trials.

Glycogen synthetic kinase 3β (GSK3β) is the upstream protein of tau phosphorylation and has a unique position in the development of tau pathology. At present, many small molecule inhibitors targeting GSK3β have entered clinical research.[13] GSK3β is a key target for regulating the phosphorylation of the tau protein, so many developed compounds work by inhibiting the activity of GSK3β, and they are divided into three categories. The first-generation GSK3β inhibitor was lithium. In vitro studies have shown that lithium can directly bind to GSK3β and inhibit the phosphorylation of Ser9 site, but it was found that lithium does not significantly affect cognitive ability and lacks efficacy in clinical trials.[14] The second-generation ATP-competitive inhibitors of GSK3β, including indirubin, SB415286, SB216763, and AR-A014418, have not entered clinical trials. The third-generation GSK3β non-competitive ATP-binding site inhibitors include GSK3β inhibitor 2, TDZD-8 and Tideglusib, and so forth. Tideglusib is the only third-generation GSK3β inhibitor currently in clinical trials developed by Noscira, with a heterocyclic thiadiazolidinone (TDZD) scaffold. Tideglusib showed a trend to reduce GSK3β activity in the phase IIa clinical trial, but it did not show a significant clinical effect due to the absence of its primary endpoint and some secondary endpoints in the phase IIb clinical trial (Figure ).[15]

Figure 3

ATP or non-ATP competitive inhibitor of GSK3β.

Early studies proposed that the hypotheses of Aβ, tau, and cholinergic pathology were carried out on different timelines. It has recently been found that they are not completely independent and will eventually complement each other and influence each other.[16] On one hand, Aβ participates in the process of tau protein hyperphosphorylation. Aβ can not only activate the phosphorylation site of tau, but also induce the redistribution of tau in neurons, leading to the neuronal structure and dysfunction. On the other hand, the activation of GSK3β can also depolymerize tubulin, affect axonal transport function, and promote the expression of AChE, which greatly reduces the level of acetylcholine (ACh) and affects cognitive function (Figure ).[17] In this way, triple inhibition of AChE, BACE1, and GSK3β and inhibition of cholinergic, Aβ, and tau pathological process can be proposed as a promising and preventive strategy against AD.

Figure 4

Interaction mechanism of BACE1, GSK3β, and AChE.

Our previous study found that the Notopterygium incisum extract (NRE) can improve the cognitive ability of APP/PS1 transgenic mice.[1] We first reported notopterol as a natural BACE1 and GSK3β dual inhibitor from N. incisum can reduce Aβ and phosphorylated tau and effectively improve the pathology of cognitive impairment in APP/PS1 AD mice.[18] In order to simultaneously inhibit the three classical targets of AD, we designed a series of triple inhibitors of AChE/BACE1/GSK3β lead compounds based on a furacoumarin scaffold of notopterol in this study. The pharmacodynamics (PD) of the lead compounds was investigated by the established Aβ42-induced AD mice model. Furthermore, the oral safety, oral bioavailability, and blood–brain barrier (BBB) passage of the lead compound were also evaluated.

Results and Discussion

Design of Novel Notopterol Derivers from the BACE1–GSK3β Dual Inhibitor to Triple AChE–BACE1–GSK3β Inhibitors

Previous studies used fragment-based drug design methods to obtain a series of small molecules with good affinity and pharmacokinetic (PK) characteristics for two targets.[19] In another study, using the versatility of curcumin scaffolds, a series of BACE1 and GSK3β dual target inhibitor molecules were designed using a structure-based method.[20] In addition, in order to find fragments that are active on both BACE1 and GSK3β, some studies have conducted a virtual screening scheme and found that 1,7-dihydro-2H-pyrrolo[2,3-d]pyrimidine-2,4(3H)-diketones have weak inhibitory effects on these two targets.[3] To analyze the interaction of notopterol with BACE1 and GSK3β, we performed docking analysis and found that notopterol can effectively bind to these two targets. The hydroxyl group at the end of the flexible chain of notopterol forms a key hydrogen bond with Asp32 of BACE1, while the lactone ring of the scaffold forms key hydrogen bonds with Tyr134 and Val135 of GSK3β, as shown in Figure . These results suggested that the furanocoumarin scaffold and aliphatic chain were necessary groups for the dual inhibition of BACE1 and GSK3β. Since it is large enough for structural modification of the aliphatic chain of N, this study intends to preserve the furanocoumarin scaffold and modify the structure of the flexible chain. Therefore, in this study, aliphatic chains of different chain lengths were designed and replaced with terminal N in order to enhance their inhibitory activity on BACE1.

Figure 5

Molecular docking of notopterol with BACE1 and GSK3β.

In order to enhance the inhibitory activity of notopterol derivatives on AChE, we carried out the following design. The compound design strategies in this study use the furanocoumarin moiety in notopterol as the scaffold, the propyl chain as the linker, and connected the 4-aminopiperidine ring with BACE1 inhibitory ability.[21] The biological activities of these compounds were evaluated in order to screen out the triple inhibitor of AchE, BACE1, and GSK3β as leading compounds (Figure ).

Figure 6

Design strategy for triple-targeted inhibitors of AchE, BACE1, and GSK3β.

Chemistry

We designed three routes to investigate the length of the aliphatic chain and the number of N atoms, as well as the influence of the intermediate steric hindrance and the aromatic group at the end of the aliphatic chain. The synthesis of notopterol derivatives was based on bergapten as the starting material. According to the previous method,[22] bergapten was demethylated by BBr3 to obtain bergaptol (2). Then, 2 undergoes a halogenation reaction or acylation reaction and goes through different routes to obtain compounds A1–A9, B1–B8, and C1–C11.

First, we designed compounds A1–A4 with a primary amine at the end of different chain lengths, as shown in Scheme . N-boc-bromoethylamine, N-boc-bromopropylamine, and N-boc-bromobutylamine were added to halogenate with 2 to obtain A1, A2 and A3, respectively. To synthesize notopterol derivatives with different numbers of N atoms terminated in the aliphatic chain, we halogenated A2 with N-Boc-bromoethylamine or N-Boc-bromopropylamine and added ethyl acetate hydrochloride to remove the Boc group to obtain A5 and A6. In order to investigate the relationship between the alkalinity of N at the end of the aliphatic chain, we designed and synthesized notopterol derivatives with amide fragments at the end. 3-Chloropropionamide was obtained by the reaction of 3-chloropropionyl chloride with ammonia and then halogenated with 8 to finally obtain A7. A similar method was obtained for A8.

Scheme 1

Reagents and Conditions: (a) BBr3, DCM, 0 °C, 4 h. (b) A1–A6

(1) The corresponding base, acetone, K2CO3, NaI, 60 °C, reflux. (2) 4 M HCl–EtOAc, 2 h, rt. A7–A9: 3-chloropropionyl chloride/4-chlorobutyryl chloride acetone/5-chlorovaleryl chloride, THF, NH3·H2O. (3) K2CO3, NaI, 60 °C, reflux.

As to limit the flexibility of the front segment of the aliphatic chain, we designed Scheme and synthesized a total of 8 compounds. 3-Chloropropionyl chloride and 8 were acylated to obtain intermediate B1-1, and then the product was reacted with Boc-methylamino to obtain B1. B2 and B3 were obtained by a similar method. Next, we attached an aromatic group to the end of the derivative aliphatic chain. B4 and B5 were obtained by halogenation with bromobenzyl on the basis of B2 and B3. B6, B7, and B8 were obtained by a similar method.

Scheme 2

Reagents and Conditions: (a) 3-Chloropropionyl Chloride/4-chlorobutyryl Chloride, Et3N, DCM, rt, 6 h; (b) Corresponding Base, Acetone, K2CO3, NaI, 60 °C, Reflux

According to the inhibitory activity of the above synthesized derivatives on BACE1, it was found that the length of the aliphatic chain should not be too long, and n = 3 was appropriate (A2). Therefore, we fixed the length of the aliphatic chain in Scheme , connected various rigid groups at the end by the halogenation reaction, and obtained C1–C11.

Scheme 3

Reagents and Conditions: (a) Corresponding Base, Acetone, K2CO3, NaI, 60 °C, Reflux; (b) 4 M HCl–EtOAc, 2 h, rt

In the next study, we are committed to the design and synthesis of notopterol derivatives with triple BACE1–GSK3β–AChE inhibition. Bergapten was selected as the starting material. First, bergaptol (2) was obtained by the reaction of boron tribromide under the protection of nitrogen and then reacted with 1,3-dibromopropane and 4-tert butoxycarbonyl aminopiperidine for two nucleophilic substitution reactions. The tert butoxycarbonyl was removed by trifluoroacetic acid to form trifluoroacetate, and then the key intermediates 4–6 were obtained. Finally, notopterol derivatives 1a–3e were obtained by reductive amination (Scheme ).

Scheme 4

Synthesis of 1a–3e

Reagents and conditions: (a) BBr3, DCM, 0 °C, 4 h; (b) 1,3-dibromopropane, acetone, K2CO3, NaI, reflux, 12 h; (c) (1) 4-Boc-aminopiperidine, acetone, K2CO3, NaI, reflux. (2) CF3COOH, rt, 4 h; Et3N, pH = 9–10; (d) (1) Correspoding benzaldehyde, methanol, acetic acid, reflux. (2) 0 °C, sodium cyanoborohydride, 1 h.

Next, in order to investigate the effect of the position of aminopiperidine, we designed and synthesized the following notopterol derivatives. After the nucleophilic substitution reaction between 4-tert butoxycarbonyl aminopiperidine and the corresponding F-substituted benzyl, tert butoxycarbonyl was removed to form trifluoroacetate, and triethylamine was added to obtain intermediate 7. Then, it was halogenated with 3 to obtain notopterol derivatives 4a–4e (Scheme ).

Scheme 5

Synthesis of 4a–4e

Reagents and conditions: (a) (1) acetone, K2CO3, NaI, reflux, 12 h. (2) CF3COOH, rt, 4 h; (b) acetone, K2CO3, NaI, reflux.

Biological Evaluation

AChE/BACE1/GSK3β Inhibitory Activity

Next, the inhibitory activity of the above-mentioned compounds on BACE1 and GSK3β were tested in this study (Table ). Most of the notopterol derivatives showed moderated inhibitory activity against BACE1 and GSK3β, as shown in Table . When the length of the linker is propyl, the inhibitory activity on BACE1 is better, like the BACE1 inhibition rate of A2 was 54.7% (20 μM). After fixing the linker length, the inhibitory activities of different groups on BACE1 were investigated. When the amino terminal was linked, C11 with aminopiperidine showed the strongest inhibition rate (BACE1 inhibition was 62.3%). However, the inhibitory activity of these fragments on GSK3β is not significantly different.

Table 1

Inhibitory Activity of the Compound on BACE1 and GSK3β

compound	inhibitiona(%, hBACE1)	inhibitionb(%, hGSK3β)	compound	inhibitiona(%, hBACE1)	inhibitionb(%, hGSK3β)
A1	34 ± 1	35 ± 1	B6	13 ± 1	37 ± 1
A2	55 ± 1	34 ± 1	B7	60 ± 1	34 ± 1
A3	30 ± 1	36 ± 1	B8	33 ± 2	34 ± 1
A4	38 ± 1	35 ± 1	C1	36 ± 1	38 ± 1
A5	24 ± 1	36 ± 1	C2	33 ± 1	39 ± 1
A6	12 ± 1	36 ± 1	C3	46 ± 1	38 ± 1
A7	46 ± 1	33 ± 1	C4	40 ± 1	38 ± 1
A8	46 ± 1	37 ± 1	C5	26 ± 1	40 ± 1
A9	33 ± 1	36 ± 1	C6	42 ± 1	38 ± 1
B1	34 ± 1	34 ± 1	C7	31 ± 1	39 ± 1
B2	29 ± 1	35 ± 1	C8	40 ± 1	38 ± 1
B3	34 ± 1	35 ± 1	C9	40 ± 1	40 ± 1
B4	34 ± 1	34 ± 1	C10	41 ± 1	40 ± 1
B5	54 ± 1	37 ± 1	C11	62 ± 1	44 ± 1
notopterolc	45 ± 1	30 ± 1	LY2811376d	52 ± 1	inactive
staurosporinee	inactive	62 ± 1

Inhibition percentage of hBACE1 at 20 μM.

Inhibition percentage of hGSK3β at 10 μM.

The concentration of notopterol for hBACE1 is 20 μM, and for hGSK3β is 10 μM.

The concentration of LY2811376 is 5.0 μM.

The concentration of staurosporine is 1.0 μM.

It can be seen from Table that most of the notopterol derivatives designed and synthesized in this study have good inhibitory activities on AChE, BACE1, and GSK3β. The inhibitory activities of most derivatives on BACE1 were similar, and the inhibitory rates of 2c and 4c on BACE1 were 49.5 and 49.3%, respectively. These results suggested that the position of aminopiperidine had no important effect on the inhibitory activity of BACE1. Similarly, there was no significant difference in the inhibitory activity of these derivatives on GSK3β, among which 4d and 1c with the best activity had 42.0 and 40.3% inhibition rates of GSK3β, respectively. Unexpectedly, the inhibitory activities of these compounds against AChE were quite different. Among these compounds, 1c exhibited the strongest inhibitory activity against AChE at 58.7%. Taken together, 1c showed balanced inhibitory effects on BACE1 (IC50 = 20 ± 1 μM), GSK3β (IC50 = 15 ± 1 μM), and AChE (IC50 = 1 ± 0 μM). As shown in the Table , 1c had a strong inhibitory effect on AChE.

Table 2

Inhibitory Activity of Notopterol Derivatives on BACE1, GSK3β, and AChE

compound	F position	inhibition (%, hBACE1)a	inhibition (%, hGSK3β)b	inhibition (%, AChE)c
1a		48 ± 1	35 ± 1	38 ± 2
1b	4	45 ± 1	36 ± 1	52 ± 1
1c	3	48 ± 1	40 ± 1	59 ± 1
1d	3, 4	49 ± 1	34 ± 1	37 ± 1
1e	2, 4	40 ± 1	25 ± 1	44 ± 1
2a		40 ± 1	36 ± 1	36 ± 1
2b	4	44 ± 1	38 ± 1	46 ± 2
2c	3	50 ± 1	40 ± 1	47 ± 1
2d	3, 4	44 ± 1	37 ± 1	38 ± 2
2e	2, 4	38 ± 1	30 ± 1	39 ± 1
3a		42 ± 2	33 ± 1	39 ± 1
3b	4	44 ± 1	36 ± 1	30 ± 1
3c	3	30 ± 1	40 ± 1	20 ± 1
3d	3, 4	32 ± 1	20 ± 1	19 ± 1
3e	2, 4	29 ± 1	39 ± 1	19 ± 1
4a		44 ± 1	30 ± 1	10 ± 1
4b	4	39 ± 2	34 ± 1	29 ± 1
4c	3	49 ± 1	33 ± 1	19 ± 1
4d	3, 4	30 ± 1	42 ± 1	15 ± 1
4e	2, 4	40 ± 1	39 ± 1	11 ± 1
LY2811376d		56 ± 1	inactive	inactive
staurosporinee		inactive	60 ± 1	inactive
tacrinef		inactive	inactive	69 ± 1
Notopterol		35 ± 1	30 ± 1	inactive

Inhibition percentage of hBACE1 at 20 μM.

Inhibition percentage of hGSK3β at 10 μM.

Inhibition percentage of AChE at 1.0 μM.

The concentration of LY2811376 is 5.0 μM.

The concentration of staurosporine is 1.0 μM.

The concentration of tacrine is 1.0 μM.

Table 3

IC50 of AChE, BACE1, and GSK3β of Compound 1c

	IC50
compounds	AChE	BACE1	GSK3β
1c	1 ± 0.4 μM	20 ± 1 μM	15 ± 1 μM
LY2811376		4 ± 0.03 μM
staurosporine			0.03 ± 0.001 μM
tacrine	0.05 ± 0.008 μM

Brain Permeation In Vitro

The permeability of the BBB is a very important property of the lead compound of central nervous drugs. In this study, a model of passive transcellular permeation in vitro was reported from our previous study,[23] along with parallel artificial membrane permeation test BBB (PAMPA-BBB), which was used to evaluate the BBB permeability of 1c. The determination method was verified by comparing the experimental permeability (Pe) values and reported permeability (Pe) values of 10 commercially available drugs (Table ). According to the obtained linear correction Pe (exp.) = 0.6984 Pe (lit.) + 0.6369 (R2 = 0.9796) and the limit determined by Di et al.,[24] we conclude that a compound with the permeability greater than 4.7 × 10–6 cm/s could pass through BBB (CNS+). For low (CNS−) and uncertain (CNS±) BBB penetration, the thresholds of Pe < 3.2 and 4.7 > Pe > 3.2 were established, respectively. These results showed that 1c could pass through BBB (Pe > 8.1 × 10–6).

Table 4

Permeability (Pe) of Commercial Drugs and 1c in the PAMPA-BBB Assay

commercial drugs	lit.a	exp.b	cpd	Pe(10–6 cm/s)	prediction
enoxacin	0.9	0.8	1c	8	CNS+
ofloxacin	0.8	0.8
verapamil	16	12
hydrocortisone	2	2
caffeine	1	3
progesterone	9	7
testosterone	17	12
piroxicam	2	2
atenolol	0.8	1
clonidine	5	4

Taken from refs (25) and (26).

Data are the mean of the two independent experiments.

Docking and Molecular Dynamics Simulation

To gain better insights into the possible interaction of 1c with AChE, BACE1, and GSK3β, molecular docking studies were performed by using Glide implemented in Schrödinger software. As shown in Figure A, the furocoumarin scaffold of 1c occupies the CAS site of AChE and binds to the benzene ring of Trp86 through π–π stacking, and the oxygen on the furan ring generates hydrogen bonding with Ser125. The intermediate piperidine formed a hydrophobic interaction with Tyr341, while the benzyl group occupies the PAS site and hydrophobic interaction with Ser293. Figure C can be accommodated in the S2′, S1, and S3 active cavities of BACE1, the benzene and furan rings of the furocoumarin scaffold of 1c generate π–π stacking interactions with Ile110 and Thr231 of BACE1, the nitrogen atoms connected to the piperidine ring can attract the charge of the key amino acids Asp32 and Asp228 to generate hydrogen bonds, and the terminal benzyl group has a hydrophobic interaction with Tyr198. The formation of the active pocket of 1c and GSK3β had an important hydrophobic interaction, that is, the furan ring in the furocoumarin scaffold had a hydrophobic interaction with Gly63, and the benzene ring of the benzyl group had a hydrophobic interaction with Ala83, Val110, and Leu188 to generate π-alkyl. The secondary amine formed a key hydrogen bond with Val135 (Figure E).

Figure 7

Molecular docking and MD simulation of 1c with AChE, BACE1, and GSK3β. (A) Predicted binding modes of 1c and AChE (PDB: 4EY7). (B) rmsd of the 1c-AChE complex within 100 ns. (C) Predicted binding modes of 1c and BACE1 (PDB: 5CLM). (D) rmsd of the 1c-BACE1 complex within 100 ns. (E) Predicted binding modes of 1c and GSK3β (PDB: 4PTC). (F) rmsd of the 1c-GSK3β complex within 100 ns. Docking was performed with Glide, and images were generated with Pymol. Green color represents hydrogen bonding interactions, and pink color represents hydrophobic interactions.

In addition, we performed molecular dynamics (MD) simulations to investigate the stability of the docking complexes of 1c against AChE, BACE1, and GSK3β. The protein root mean square deviation (rmsd) values were monitored for each simulation run to measure the stability of protein–ligand poses. In this study, the protein rmsd values of 1c-AChE, 1c-BACE1, and 1c-GSK3β complexes were within the acceptable range of 0–3 Å during the 100 ns simulation. The conformational change of the docking ligand rmsd complex was within 0–3 Å (within 10 Å is acceptable) (Figure B,D,F).

Safety Profile of 1c in C57B6/J Mice

In order to explore the oral safety of 1c, we investigated the acute toxicity of 1c. For single-dose administration, we investigated the effects of 400 and 200 mg/kg 1c on the liver and kidney function in mice, respectively. As illustrated in Figure A,B, after a single dose of 400 and 200 mg/kg to mice, serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) of C57B6/J mice increased slightly, but there was no significant difference compared with the control group. Similarly, there was no significant change in the content of blood urea nitrogen (BUN) in mice treated with 1c (Figure C). Furthermore, the liver and kidney of mice were examined pathologically. As shown in Figure D, the morphology of liver and kidney tissue of 1c-exposed mice did not change significantly.

Figure 8

Acute toxicity of 1c in C57BL/6 mice. (A) Content of ALT in mice serum, n = 9. (B) AST content in mice serum, n = 9. (C) BUN content in mice serum, n = 9. (D) H&E staining of the liver and kidney of mice, scale bar: 50 μm. The error bars represent the SD.

PK Properties of 1c

The PK profile of 1c was investigated before the PD investigation. We obtained PK profiles through oral (p.o., 100 mg/kg) and intravenous (i.v., 10 mg/kg) to male Sprague–Dawley (SD) rats (shown in Table ). After a single 10 mg/kg i.v. dose of 1c, Cmax was 2796 ± 259 ng/mL, AUC0– was 1032 ± 86 μg/mL*h, and t1/2 was 0.4 ± 0.04 h. At a dose of 100 mg/kg (oral) 1c, Cmax was 167 ± 13 ng/mL, AUC0– was 1010 ± 112 μg/mL*h, and t1/2 was 20 ± 9 h. The oral bioavailability of 1c was 9.8%.

Table 5

PK Profile of 1ca

parameters	100 mg/kg (p.o.)	10 mg/kg(i.v.)
Cmax(ng/mL)	167 ± 13	2796 ± 259
AUC0–t(ng/mL)	1010 ± 112	1031 ± 86
AUC0–∞(ng/mL*h)	1635 ± 362	1047 ± 88
t1/2 (h)	20 ± 9	0.4 ± 0.04
Cl (L/h/kg)	63 ± 12	10 ± 1
MRT0–∞ (h)	26 ± 11	0.3 ± 0
VZ(L/kg)	1730 ± 387	5 ± 1
Tmax (h)	1	0.08
F (%)	9.8

PK parameters (mean ± SD, n = 5); Cmax, maximum drug concentration; AUC, area-under-the-curve; Cl, plasma clearance rate; VZ, steady state volume of distribution; MRT, mean residence time; t1/2, terminal half-life, Tmax, the time take to reach Cmax; F, oral bioavailability; I.V., intravenous administration; P.O., oral administration.

In Vivo PD Study

We used a mouse brain stereotaxic instrument to inject Aβ42 into the lateral ventricle of mice to cause memory and cognitive dysfunction in mice, mimicking the symptoms of AD. As shown in Figure A, the directional navigation results showed that the escape latency of the mice injected with Aβ was significantly longer than that of the control group, and there was no significant difference in the swimming speed of the mice in each group (Figure B). The escape latency of mice given 1c was less than that of the model group (Figure C). The results of spatial exploration showed that the number of crossing platforms in mice given 1c was significantly higher than that in the Aβ group, and there was a dose-dependent trend (Figure D,E). These results indicated that oral administration of 1c can significantly improve the cognitive impairment of AD mice.

Figure 9

Morris water maze (MWM). (A) Representative trajectory of directional navigation. (B) Swimming speed and escape latency (C) of each group mice. (D) Representative trajectory of space exploration. (E) nNumber of times mice in each group crossed the platform. The error bars represent the SD, n = 8.

Effects of 1c on the Expression of Aβ-related Proteins

Given the effect of 1c on cognitive impairment in Aβ-induced AD mice, we examined the expression of Aβ-related proteins in cortical and hippocampal tissues. As illustrated in Figure A–C, the expression of ADAM17 and BACE1 in Aβ mice was significantly increased compared to controls, and 1c can inhibit the expression of ADAM17 in the cortex. 1c had an inhibitory effect on cortical BACE1 but not statistically significant. We also examined the expression levels of ADAM17 and BACE1 in the hippocampus. Similarly, ADAM17 and BACE1 expression in the hippocampus of Aβ mice was significantly increased compared to controls (Figure D). The expressions of ADAM17 and BACE1 were significantly decreased in AD mice after 1c treatment (Figure E,F).

Figure 10

ADAM17 and BACE1 expression in the cortex and hippocampus. (A) Immunoblot bands for ADAM17 and BACE1 in the mouse cortex. Statistical histogram of ADAM17 (B) and BACE1 (C) expression in the mouse cortex. (D) Immunoblot bands for ADAM17 and BACE1 in the mouse hippocampus. Statistical histogram of ADAM17 (E) and BACE1 (F) expression in the mouse hippocampus. The error bars represent the SD, n = 3.

Conclusions

In conclusion, this study synthesized a series of novel notopetrol derivatives and evaluated their inhibitory effects on AChE, BACE1, and GSK3β in enzymatic assays. Among them, 1c showed a balanced inhibitory activity against the three. In addition, the PAMPA-BBB results indicated that 1c could penetrate the blood–brain barrier. Molecular docking results showed that 1c could interact with key amino acids of AChE, BACE1, and GSK3β. MD simulation experiments show that 1c can form stable conformations with AChE, BACE1, and GSK3β within 100 ns. In addition, 1c has a good oral safety profile, but the bioavailability needs to be improved, and the chemical structure needs to be further improved. Importantly, 1c effectively ameliorated the cognitive impairment of Aβ-induced AD mice and attenuated the expression levels of Aβ-related proteins in the cortex and hippocampus of AD mice in vivo. The specific mechanism of 1c regulating the balance of AChE, BACE 1, and GSK 3β targets needs to be further studied in vivo and vitro. These results suggest that 1c is a potential multi-target lead compound that can be further structurally modified to treat AD.

Experimental Protocols

General Procedures

As described in our previous study,[23] all reagents and solvents were purchased from commercial supplies and used at the highest available purity without further purification. All reactions involving air or moisture sensitive intermediates were carried out under nitrogen. All target compounds were purified by silica gel column chromatography. 1H NMR and 13C NMR spectra were recorded as internal standards in CDCl3 or DMSO-d6 using the Bruker AVIII-600 (Bruker Corporation, Bremen, Germany) at 1H at 600 MHz and 13C at 150 MHz. Chemical shifts (δ) are reported in parts per million (ppm) using tetramethylsilane as an internal standard. The coupling constant J is expressed in Hertz (Hz). The HR-TOFMS was measured on the Bruker micro TOFQ mass spectrometer system. Column chromatography was performed on silica gel (200–300 mesh) from Qingdao Ocean Chemical (Qingdao, China). Thin-layer chromatography (TLC) was performed on 20 mm precoated silica gel plates (Merck, Silica 60 F254); visualization was achieved using UV light (254 nm).

Synthesis of 4-Hydroxy-7H-furo[3,2-g]chromen-7-one (2)

Bergapten (10 g, 46.2 mmol) was taken and added to pre-cooled dichloromethane (100 mL), and BBr3 (46.2 mL, 46.2 mmol, 1 M in DCM) was slowly added under nitrogen and stired at 0 °C for 4 h. The mixed solution was slowly poured into ice water, and a dark yellow solid was precipitated. After the solid was cooled, the sample was recovered by filtration and dried to obtain an off-white solid, yield 95%. ESI-MS: m/z 203.1 [M + H]+.

Synthesis of 4-(3-Bromopropoxy)-7H-furo[3,2-g]chromen-7-one (3)

Compound 2 (8.3 g, 25.7 mmol), 4-tert-butoxycarbonylaminopiperidine (20.8 g, 102.5 mmol), K2CO3 (7.1 g, 51.3 mmol), and NaI (2.1 g, 12.8 mmol) were added to a 250 mL round-bottom flask, and 100 mL of acetone was added as a solvent and refluxed at 60 °C overnight. After the reaction was monitored by TLC, the reaction was terminated, the solid was removed by suction filtration, purified by silica gel column chromatography, eluted with the mobile phase of petroleum ether and ethyl acetate (10:1 to 1:1), and a yellow oil was obtained. The oily product was transferred to a 100 mL round-bottom flask, 20 mL of trifluoroacetic acid and 20 mL of dichloromethane were added, and the mixture was stirred at room temperature for 2 h. After monitoring by TLC until the reaction was complete, the reaction solution was evaporated to dryness under vacuum. Then, 20 mL of dichloromethane was added, triethylamine was gradually added in an ice bath until the pH value reaches 9–10, stirred at room temperature for 30 min, and the solvent was evaporated to dryness. Then, it was extracted with ethyl acetate and water, the ethyl acetate extract was recovered, evaporated to dryness under reduced pressure, and purified by silica gel column chromatography. The mobile phase was eluted with dichloromethane/methanol (50:1 to 10:1), 0.1% triethylamine was added to the mobile phase to obtain fractions of the target compound, and the solvent was removed to obtain a pale yellow solid with a yield of 67%.

Synthesis of 2-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)ethan-1-aminium Chloride (A1)

8 (100 mg, 0.495 mmol), N-Boc-bromoethylamine (1.1 g, 4.95 mmol), 10 mL of acetone, K2CO3 (136.6 mg, 0.99 mmol), and NaI (36.75) were added to a 50 mL eggplant-shaped flask (mg, 0.24 mmol), heated to 60 °C, and refluxed overnight. TLC detection showed that there are still raw materials remaining. Purification by column chromatography furnished the desired compound after elution with petroleum ether/ethyl acetate (20:1 to 10:1), and the related fractions were combined and removed under reduced pressure, solvated to obtain a light yellow oil, and transferred to a 50 mL eggplant-shaped flask. 8 mL of hydrochloric acid ethyl acetate was added and stirred at room temperature for 2 h; a white solid precipitated out and was filtered and dried to obtain a pale yellow solid with a yield of 63%. 1H NMR (400 MHz, DMSO-d6): δ 8.56 (1H, d, J = 12.0 Hz), 8.10 (1H, d, J = 4.0 Hz), 7.44 (1H, s), 7.38 (1H, d, J = 4.0 Hz), 6.40 (1H, d, J = 12.0 Hz), 4.69 (2H, t, J = 8.0, 12.0 Hz), 3.79 (2H, s). 13C NMR (100 MHz, DMSO-d6): δ 160.6, 158.0, 152.6, 148.3, 146.8, 140.8, 113.4, 112.7, 106.6, 105.8, 94.4, 69.5, 39.3. HR-ESI-MS: m/z 246.0697 [M + H]+ calcd for C13H11NO4, 245.0688.

Synthesis of 3-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)propan-1-aminium Chloride (A2)

Yield 73%, white solid. The synthesis method is the same as that of A1, and the raw material is replaced with N-Boc-bromopropylamine. 1H NMR (400 MHz, DMSO-d6): δ 8.29 (1H, d, J = 12.0 Hz), 8.06 (1H, d, J = 4.0 Hz), 7.42 (1H, d, J = 4.0 Hz), 7.37 (1H, s), 6.34 (1H, d, J = 12.0 Hz), 4.63 (2H, t, J = 8, 12.0 Hz), 3.43 (2H, t, J = 8, 12.0 Hz), 3.04 (2H, d, J = 4.0 Hz), 2.16 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.6, 148.9, 146.5, 140.3, 113.4, 112.8, 106.5, 106.1, 94.0, 72.3, 53.4, 38.5. HR-ESI-MS: m/z 260.0873 [M + H]+ calcd for C14H13NO4, 259.0845.

Synthesis of 3-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)butan-1-aminium Chloride (A3)

Yield 54%, white solid. The synthesis method is the same as that of A1, and the raw material is replaced with N-Boc-bromobutylamine. 1H NMR (400 MHz, MeOD): δ 8.38 (1H, d, J = 12.0 Hz), 7.81 (1H, d, J = 4.0 Hz), 7.27 (1H, d, J = 4.0 Hz), 7.19 (1H, s), 6.32 (1H, d, J = 12.0 Hz), 4.63 (2H, t, J = 8, 12.0 Hz), 3.42 (2H, t, J = 8, 12.0 Hz), 3.00 (2H, d, J = 8.0 Hz), 2.16 (2H, m). 13C NMR (150 MHz, MeOD): δ 160.6, 158.1, 152.6, 149.1, 146.5, 140.8, 113.5, 112.9, 106.5, 106.1, 93.8, 72.4, 39.0, 26.8, 24.2. HR-ESI-MS: m/z 274.1034 [M + H]+ calcd for C15H15NO4, 273.1001.

Synthesis of 3-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)pentan-1-aminium Chloride (A4)

Yield 35%, white solid. The synthesis method is the same as that of A1, and the raw material is replaced with N-Boc-bromopentylamine. 1H NMR (400 MHz, MeOD): δ 8.38 (1H, d, J = 12.0 Hz), 7.81 (1H, d, J = 4.0 Hz), 7.27 (1H, d, J = 4.0 Hz), 7.19 (1H, s), 6.32 (1H, d, J = 12.0 Hz), 4.64 (2H, t, J = 8.0, 12.0 Hz), 3.61 (2H, d, J = 8.0 Hz), 3.00 (2H, t, J = 8, 16.0 Hz), 2.24 (2H, m), 1.32 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.0, 152.4, 149.1, 146.3, 140.1, 113.2, 112.5, 106.3, 106.0, 93.4, 70.8, 37.0, 30.4, 28.6, 21.4. HR-ESI-MS: m/z 288.1173 [M + H]+ calcd for C16H17NO4, 287.1158.

Synthesis of 4-(3-((2-(l4-Azanyl)ethyl)amino)propoxy)-7H-furo[3,2-g]chromen-7-one, Chloride Salt (A5)

Yield 32%, white solid. A2 (100 mg, 0.495 mmol), N-Boc-bromoethylamine (740 mg, 4.95 mmol), 10 mL of acetone, K2CO3 (273.2 mg, 1.98 mmol), and NaI (36.75 mg, 0.24 mmol) were added, heated to 60 °C, and refluxed overnight. After TLC detection, it was found that the reaction was complete, and then the solvent was removed under reduced pressure to obtain a light yellow oil, which was transferred to a 50 mL eggplant-shaped flask, and 8 mL of hydrochloric acid ethyl acetate was added. The mixture was stirred at room temperature for 2 h. A white solid precipitated out and was dried by suction. The final product is then obtained. 1H NMR (600 MHz, DMSO-d6): δ 8.29 (1H, d, J = 12.0 Hz), 8.05 (1H, d, J = 6.0 Hz), 7.37 (2H, m), 6.32 (1H, d, J = 12.0 Hz), 4.63 (2H, t, J = 6, 12.0 Hz), 3.17 (2H, d, J = 6.0 Hz), 3.10 (2H, d, J = 6.0 Hz), 1.82 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 168.9, 153.4, 152.6, 149.0, 146.4, 140.2, 120.6, 118.6, 116.6, 114.7, 93.6, 69.9, 47.4, 45.9, 36.5, 21.4. HR-ESI-MS: m/z 289.1118 [M + H]+ calcd for C15H16N2O4, 288.1110.

Synthesis of 3-((3-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)propyl)amino)propan-1-aminium Chloride (A6)

Yield 28%, light white solid. The synthesis method is the same as that of A5, and the raw material is replaced with N-Boc-bromopropylamine. 1H NMR (600 MHz, DMSO-d6): δ 8.29 (1H, d, J = 12.0 Hz), 8.05 (1H, d, J = 6.0 Hz), 7.37 (2H, m), 6.32 (1H, d, J = 12.0 Hz), 4.63 (2H, t, J = 6, 12.0 Hz), 4.16 (2H, t, J = 6, 12.0 Hz), 3.17 (2H, d, J = 6.0 Hz), 3.10 (2H, d, J = 6.0 Hz), 1.82 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.7, 153.5, 152.6, 149.0, 146.4, 140.2, 120.6, 118.6, 116.6, 114.7, 93.7, 66.7, 53.7, 36.8, 29.4, 28.0, 21.4. HR-ESI-MS: m/z 303.1287 [M + H]+ calcd for C16H18N2O4, 302.1267.

Synthesis of 3-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)propanamide (A7)

1 mL of 3-chloropropionyl chloride was taken and dissolved in 10 mL of THF, then the solution was slowly added dropwise to pre-cooled ammonia (13.3 mol/L), andstirring was continued for 4 h. After the reaction was completed, it was extracted with dichloromethane, and the solvent was rotary evaporated to obtain a yellow oily intermediate, namely, 3-chloropropionamide. Then100 mg of 8 (0.495 mmol), 3-chloropropionamide (1.98 mmol, 212.9 mg), 10 mL of acetone, K2CO3 (136.6 mg, 0.99 mmol), and NaI (36.75 mg, 0.24 mmol) were taken, heated to 60 °C, and refluxed overnight. Purification by column chromatography furnished the desired compound after elution with petroleum ether/ethyl acetate (20:1 to 1:1). The related fractions were combined, and the solvent was removed under reduced pressure to obtain a pale yellow oil, yield 38%. 1H NMR (600 MHz, DMSO-d6): δ 8.25 (1H, d, J = 12.0 Hz), 7.91 (1H, d, J = 6.0 Hz), 7.22 (2H, d, J = 6.0 Hz), 7.15 (1H, s), 6.25 (1H, d, J = 12.0 Hz), 4.26 (2H, m), 3.08 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 173.8, 160.9, 157.5, 153.2, 148.5, 145.4, 140.4, 113.0, 111.4, 105.3, 104.2, 91.4, 60.8, 40.0. HR-ESI-MS: m/z 274.0630 [M + H]+ calcd for C14H11NO5, 273.0637.

Synthesis of 4-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)butanamide (A8)

Yield 32%, pale yellow oil. The synthesis method is the same as that of A8, and the raw material is replaced with 4-chlorobutyryl chloride. 1H NMR (600 MHz, DMSO-d6): δ 8.13 (1H, d, J = 12.0 Hz), 7.47 (1H, d, J = 6.0 Hz), 7.09 (2H, m), 6.24 (1H, m), 4.39 (2H, t, J = 6, 12.0 Hz), 2.65 (2H, m), 1.41 (2H, s). 13C NMR (150 MHz, DMSO-d6): δ 174.8, 169.2, 162.9, 155.6, 155.3, 143.6, 141.4, 114.9, 106.7, 106.3, 103.6, 94.2, 63.1, 46.1, 33.7. HR-ESI-MS: m/z 288.0763 [M + H]+ calcd for C15H13NO5, 287.0794.

Synthesis of 5-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)pentanamide (A9)

Yield 25%, pale yellow oil. The synthesis method is the same as that of A8, and the raw material is replaced with 5-chlorovaleryl chloride. 1H NMR (600 MHz, DMSO-d6): δ 8.13 (1H, d, J = 12.0 Hz), 7.47 (1H, d, J = 6.0. Hz), 7.09 (2H, m), 6.24 (1H, m), 3.17 (2H, m), 2.38 (2H, m), 1.96 (2H, m), 141 (2H, s). 13C NMR (150 MHz, DMSO-d6): δ 177.0, 169.2, 162.7, 155.6, 155.3, 143.6, 141.9, 114.9, 106.7, 106.3, 103.6, 94.2, 61.4, 48.1, 33.7, 23.2. HR-ESI-MS: m/z 302.0967 [M + H]+ calcd for C16H15NO5, 301.0950.

Synthesis of 7-Oxo-7H-furo[3,2-g]chromen-4-yl 3-Chloropropanoate (B1-1)

200 mg of 8 (0.99 mmol), 3-chloropropionyl chloride (380 μL, 1.98 mmol), 552 μL of anhydrous triethylamine, and 10 mL of anhydrous dichloromethane were added to a 100 mL eggplant-shaped flask. After reacting at room temperature for 6 h, a white solid precipitated out which was washed with cold methanol and dried to obtain a colorless oil, yield 89%. ESI-MS: m/z 293.6 [M + H]+.

Synthesis of 7-Oxo-7H-furo[3,2-g]chromen-4-yl 4-Chlorobutanoate (B1-2)

The synthetic method is the same as that of B1-1. Colorless oil, yield 85%. ESI-MS: m/z 307.4 [M + H]+.

Synthesis of 7-Oxo-7H-furo[3,2-g]chromen-4-yl 5-Chloropentanoate (B1-3)

The synthetic method is the same as that of B1-1. Colorless oil, yield 82%. ESI-MS: m/z 321.2 [M + H]+.

Synthesis of 7-Oxo-7H-furo[3,2-g]chromen-4-yl 3-(l4-azanyl)propanoate (B1)

100 mg of B1-1 (0.342 mmol), methylamino boc (359.4 mg, 2.74 mmol), 10 mL of acetone, K2CO3 (94.4 mg, 0.68 mmol), and NaI (23.9 mg, 0.171 mmol) were taken, heat to 60 °C, and refluxed overnight. Purification by column chromatography furnished the desired compound after elution with petroleum ether/ethyl acetate (20:1 to 1:1), the related fractions were combined, and the solvent was removed under reduced pressure to obtain a pale yellow oil, yield 23%. 1H NMR (600 MHz, DMSO-d6): δ 8.27 (1H, d, J = 12.0 Hz), 7.91 (1H, d, J = 6.0 Hz), 7.27 (1H, d, J = 6.0 Hz), 7.15 (1H, s), 6.25 (1H, d, J = 12.0 Hz), 3.55 (2H, m), 3.33 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 171.9, 160.9, 157.5, 148.4, 145.4, 140.4, 113.0, 111.4, 105.3, 104.2, 91.4, 28.8, 28.4. HR-ESI-MS: m/z 274.0642 [M + H]+ calcd for C14H11NO5, 273.0637.

Synthesis of 7-Oxo-7H-furo[3,2-g]chromen-4-yl 4-(l4-azanyl)butanoate (B2)

Yield 39%, pale yellow oil. The synthesis method is the same as that of B1, and the raw material is replaced B1-1 with 4-chlorobutyryl chloride B1-2. 1H NMR (600 MHz, DMSO-d6): δ 8.24 (1H, d, J = 12.0 Hz), 8.03 (1H, d, J = 6.0 Hz), 7.35 (5H, m), 6.44 (1H, d, J = 12.0 Hz), 4.61 (2H, t, J = 6, 12.0 Hz), 2.66 (2H, m), 2.12 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 173.8, 160.9, 157.5, 153.2, 148.5, 145.4, 140.4, 113.0, 111.4, 105.3, 104.2, 91.4, 60.8, 40.0. HR-ESI-MS: m/z 310.0798 [M + Na]+ calcd for C15H13NO5, 287.0794.

Synthesis of 7-Oxo-7H-furo[3,2-g]chromen-4-yl 5-(l4-azanyl)pentanoate (B3)

Yield 31%, pale yellow oil. The synthesis method is the same as that of B1, and the raw material is replaced B1-1 with 4-chlorobutyryl chloride B1-3. 1H NMR (600 MHz, DMSO-d6): δ 8.22 (1H, d, J = 6.0 Hz), 8.03 (1H, d, J = 6.0 Hz), 7.27 (2H, s), 7.14 (2H, s), 6.35 (1H, d, J = 12.0 Hz), 3.73 (2H, t, J = 6, 12.0 Hz), 2.66 (2H, m), 2.12 (2H, m), 2.00 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 172.5, 161.0, 160.6, 154.7, 148.5, 146.4, 145.3, 140.5, 113.7, 105.3, 104.2, 93.7, 38.0, 33.6, 31.2, 29.4. HR-ESI-MS: m/z 325.0983 [M + Na + H]2+ calcd for C16H15NO5, 301.0950.

Synthesis of 5-Oxo-5-((7-oxo-7H-furo[3,2-g]chromen-4-yl)oxy)pentan-1-aminium Chloride (B4)

B1 was used as the starting material to react with benzyl bromide, synthesized according to the above halogenation reaction conditions, and purified by silica gel column chromatography to obtain pale yellow oil, yield 37%. 1H NMR (600 MHz, DMSO-d6): δ 7.63 (1H, d, J = 12.0 Hz), 7.53 (2H, m), 7.39 (2H, d, J = 6.0 Hz), 7.32 (2H, d, J = 6.0 Hz), 7.25 (2H, m), 6.60 (1H, d, J = 12.0 Hz), 3.48 (2H, s), 2.63 (2H, m), 2.23 (2H, m), 1.88 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 172.7, 166.1, 164.6, 152.9, 143.7, 138.1, 131.8, 131.2, 131.2, 130.2, 128.6, 118.7, 116.5, 116.3, 106.7, 105.2, 93.9, 50.6, 46.4, 31.1, 29.0. HR-ESI-MS: m/z 378.3944 [M + H]+ calcd for C22H19NO5, 377.3960.

Synthesis of 7-Oxo-7H-furo[3,2-g]chromen-4-yl 5-(benzylamino)pentanoate (B5)

Pale yellow oil, yield 30%. The synthesis method is the same as that of B4, the raw material is replaced with 5-chlorovaleryl chloride. 1H NMR (600 MHz, DMSO-d6): δ 7.78 (2H, d, J = 12.0 Hz), 7.65 (1H, d, J = 12.0 Hz), 7.54 (2H, m), 7.50 (4H, m), 7.25 (2H, m), 6.68 (1H, d, J = 12.0 Hz), 3.53 (2H, s), 2.70 (2H, m), 2.33 (2H, m), 1.90 (2H, m), 1.68 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 173.1, 166.0, 162.0, 160.4, 143.6, 135.4, 134.4, 134.4, 133.5, 132.4, 130.6, 129.4, 120.0, 119.8, 119.7, 115.2, 115.0, 94.4, 50.7, 47.5, 31.0, 28.8, 24.9. HR-ESI-MS: m/z 392.4210 [M + H]+ calcd for C23H21NO5, 391.4230.

Synthesis of Methyl 1-(4-Oxo-4-((7-oxo-7H-furo[3,2-g]chromen-4-yl)oxy)butyl)piperidine-4-carboxylate (B6)

The synthesis method is the same as that of B1. According to the above halogenation reaction conditions, the raw material was replaced with methyl 4-piperidinecarboxylate, and B6 was obtained after purification by silica gel column chromatography. Yellow oil, yield 36%. 1H NMR (600 MHz, DMSO-d6): 8.29 (2H, d, J = 12.0 Hz), 7.81 (1H, d, J = 6.0 Hz), 7.24 (2H, d, J = 6.0 Hz), 6.96 (2H, s), 6.12 (1H, d, J = 12.0 Hz), 3.62 (3H, s), 2.94 (8H, m), 1.91–1.73 (5H, m). 13C NMR (150 MHz, DMSO-d6): δ 174.8, 171.8, 161.3, 157.8, 153.6, 144.5, 141.1, 113.4, 109.9, 105.8, 104.6, 89.7, 63.7, 52.5, 52.0, 45.8, 35.2, 33.7, 29.1, 28.1. HR-ESI-MS: m/z 422.1343 [M + Na]+ calcd for C21H21NO7, 399.1318.

Synthesis of Methyl 1-(5-Oxo-5-((7-oxo-7H-furo[3,2-g]chromen-4-yl)oxy)pentyl)piperidine-4-carboxylate (B7)

Yellow oil, yield 43%. B2 was used as the starting material, the synthesis method is the same as that of B6. 1H NMR (600 MHz, DMSO-d6): 7.81 (2H, m), 7.65 (2H, d, J = 12.0 Hz), 7.32 (4H, m), 7.26 (2H, m), 6.61 (1H, d, J = 12.0 Hz), 3.48 (3H, s), 2.67 (2H, m), 2.22 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 174.0, 166.1, 164.6, 162.9, 143.7, 138.9, 131.2, 131.2, 131.2, 129.2, 128.6, 127.4, 118.7, 127.4, 118.7, 118.7, 116.4, 116.3, 93.1, 50.7, 46.7, 31.1, 29.3. HR-ESI-MS: m/z 414.1493 [M + H]+ calcd for C22H23NO7, 413.1475.

Synthesis of 7-Oxo-7H-furo[3,2-g]chromen-4-yl 4-((1-benzylpiperidin-4-yl)amino)butanoate (B8)

Yellow oil, yield 21%. B2 was used as the starting material, and the raw material is replaced with 4-amino-1-benzylpiperidine. 1H NMR (600 MHz, DMSO-d6): 7.67 (4H, m), 7.39 (6H, m), 6.20 (1H, m), 3.80 (2H, s), 2.23–1.94 (10H, m), 1.12–0.94 (7H, m). 13C NMR (150 MHz, DMSO-d6): δ 174.8, 161.3, 157.8, 153.8, 144.5, 141.1, 113.4, 109.9, 105.8, 104.6, 89.9, 63.1, 56.5, 52.5, 52.0, 35.2, 33.7, 29.1, 28.1. HR-ESI-MS: m/z 461.1995 [M + H]+ calcd for C27H28N2O5, 460.1988.

Synthesis of 4-(3-Bromopropoxy)-7H-furo[3,2-g]chromen-7-one (B9)

8 (4.04 g, 20 mmol), 1,3-dibromopropane (8.12 mL, 80 mmol), K2CO3 (5.52 g, 40 mmol), and NaI (1.50 g, 10 mmol) were added into a 100 mL round bottomed flask, heated with acetone (40 mL) as the solvent, and stirred overnight at 60 °C. The recovered sample was filtered, washed with methanol, and vacuum dried to obtain a light yellow solid, yield 80%. ESI-MS: m/z 322.1 [M + H]+.

Synthesis of 4-(3-(Methylamino)propoxy)-7H-furo[3,2-g]chromen-7-one (C1)

The synthesis method is the same as that of A5, and the raw material is replaced by methylamino Boc. Purification by column chromatography furnished the desired compound after elution with petroleum ether/ethyl acetate (20:1 to 1:1), combining the related fractions and removing the solvent under reduced pressure to obtain a pale yellow oil, yield 32%. 1H NMR (400 MHz, DMSO-d6): δ 8.34 (1H, d, J = 12.0 Hz), 8.07 (1H, d, J = 4.0 Hz), 7.40 (1H, d, J = 4.0 Hz), 7.40 (1H, s), 6.35 (1H, d, J = 12.0 Hz), 4.62 (2H, t, J = 8.0, 12.0 Hz), 3.75 (2H, m), 3.38 (3H, s), 2.30 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.6, 148.9, 146.5, 140.2, 113.4, 112.8, 106.5, 106.1, 93.9, 72.3, 53.4, 38.2. HR-ESI-MS: m/z 274.1020 [M + H]+ calcd for C15H15NO4, 273.1001.

Synthesis of 4-(3-(Dimethylamino)propoxy)-7H-furo[3,2-g]chromen-7-one (C2)

The synthesis method is the same as that of A5, and the raw material is replaced by ethylamino Boc. Purification by column chromatography furnished the desired compound after elution with petroleum ether/ethyl acetate (20:1 to 1:1), the related fractions were combined, and the solvent was removed under reduced pressure to obtain a pale yellow oil, yield 24%. 1H NMR (400 MHz, DMSO-d6): δ 8.26 (1H, d, J = 12.1 Hz), 8.06 (1H, d, J = 4.0 Hz), 7.39 (1H, s), 7.34 (1H, d, J = 4.0 Hz), 6.35 (1H, d, J = 12.1 Hz), 4.55 (2H, t, J = 8.0, 12.0 Hz), 2.81 (2H, m), 2.44 (6H, s), 2.06 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.6, 149.1, 146.5, 140.1, 113.6, 112.9, 106.5, 106.0, 93.9, 70.9, 55.3, 44.5, 26.7. HR-ESI-MS: m/z 288.1166 [M + H]+ calcd for C16H17NO4, 287.1158.

Synthesis of 4-(3-(Methylamino)propoxy)-7H-furo[3,2-g]chromen-7-one (C3)

The synthesis method is the same as that of A5, and the raw material is changed to amino Boc. Pale yellow oil, yield 43%. 1H NMR (400 MHz, CDCl3-d): δ 8.16 (1H, d, J = 12.1 Hz), 7.59 (1H, d, J = 4 Hz), 7.15 (1H, s), 6.98 (1H, s), 6.28 (1H, d, J = 12.1 Hz), 4.52 (2H, t, J = 8.0, 12.1 Hz), 3.42 (2H, m), 2.09 (3H, t, J = 8.0, 12.1 Hz), 1.44 (9H, s). 13C NMR (150 MHz, CDCl3-d): δ 161.2, 158.3, 156.1, 152.7, 148.7, 144.9, 139.2, 113.3, 112.7, 106.7, 105.1, 94.1, 79.6, 70.7, 42.1, 37.7, 28.4 × 3. HR-ESI-MS: m/z 360.1357 [M + H]+ calcd for C19H21NO6, 359.1369.

Synthesis of 4-(3-Morpholinopropoxy)-7H-furo[3,2-g]chromen-7-one (C4)

The synthesis method is the same as that of A2, and the raw material is changed to morpholine. Yellow oil, yield 33%. 1H NMR (400 MHz, DMSO-d6): δ 8.38 (1H, d, J = 12.1 Hz), 7.81 (1H, d, J = 4.0 Hz), 7.26 (1H, d, J = 4.0 Hz), 7.19 (1H, s), 6.31 (1H, d, J = 12.1 Hz), 4.64 (2H, t, J = 8.0, 12.1 Hz), 3.82 (4H, m), 3.61 (2H, d, J = 4.0 Hz), 2.89 (4H, m), 2.24 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.0, 152.4, 149.1, 146.3, 140.1, 113.2, 112.5, 106.3, 106.0, 93.4, 70.8, 66.8, 56.2, 53.5, 46.5, 37.0. HR-ESI-MS: m/z 330.1251 [M + H]+ calcd for C18H19NO5, 329.1263.

Synthesis of 4-(3-(4l2-piperazin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (C5)

9 (100 mg, 0.495 mmol), 1-Boc-piperazine (740 mg, 4.95 mmol), 10 mL of acetone, K2CO3 (136.6 mg, 0.99 mmol), and NaI (36.75 mg) were added to a 50 mL eggplant-shaped flask, 0.24 mmol), heated to 60 °C, and refluxed overnight. TLC indicated that the reaction was not complete, and it was eluted with petroleum ether/ethyl acetate (20:1 to 10:1). The related fractions were combined, and the solvent was removed under reduced pressure to obtain a pale yellow oil which was transfered to a 50 mL eggplant-shaped bottle. 8 mL of ethyl acetate hydrochloride was added, stirred at room temperature for 2 h, and the white solid was precipitated, filtered, and dried to obtain a light white solid, yield 38%. 1H NMR (400 MHz, DMSO-d6): δ 8.37 (1H, d, J = 12.1 Hz), 7.80 (1H, d, J = 4.0 Hz), 7.6 (1H, d, J = 4.0 Hz), 7.19 (1H, s), 6.31 (1H, d, J = 12.0 Hz), 4.64 (2H, t, J = 8.0, 12.0 Hz), 3.98 (4H, m), 3.00 (4H, m), 2.24 (2H, m), 1.28 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 156.2, 153.5, 152.4, 149.1, 146.3, 140.1, 113.2, 112.5, 106.3, 106.0, 93.4, 76.5, 70.8, 66.8, 58.7, 55.4, 47.8, 44.0, 30.4. HR-ESI-MS: m/z 329.1467 [M + H]+ calcd for C18H20N2O4, 328.1423.

Synthesis of tert-Butyl 4-(3-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)propyl)piperazine-1-carboxylate (C6)

Light yellow oil, yield 42%. C6 is the product of C5 without de-Boc. 1H NMR (400 MHz, DMSO-d6): δ 8.26 (1H, d, J = 12.1 Hz), 8.05 (1H, d, J = 4.0 Hz), 7.37 (1H, s), 7.34 (1H, d, J = 4.0 Hz), 6.37 (1H, d, J = 12.1 Hz), 4.50 (2H, t, J = 8.0, 12.1 Hz), 3.02 (6H, m), 1.98 (3H, m), 1.72 (2H, m), 1.38 (9H, s). 13C NMR (150 MHz, CDCl3-d): δ 160.6, 158.0, 155.3, 152.6, 149.2, 146.5, 140.1, 113.8, 112.9, 106.7, 105.9, 93.9, 78.0, 71.3, 54.3, 52.5, 46.2 × 2, 31.6, 28.4 × 3, 21.2. HR-ESI-MS: m/z 429.1942 [M + H]+ calcd for C23H28N2O6, 428.1947.

Synthesis of 4-(3-(Benzylamino)propoxy)-7H-furo[3,2-g]chromen-7-one (C7)

The synthesis method is the same as that of A5, the raw material is changed to benzyl bromide, the amount of K2CO3 is reduced by half, and the reaction is found to be incomplete by TLC monitoring. Purification is performed with a silica gel column. It was eluted with petroleum ether/ethyl acetate (10:1 to 5:1), the related fractions were combined, and the solvent was removed under reduced pressure to obtain a pale yellow oil, yield 28%. 1H NMR (600 MHz, DMSO-d6): δ 7.77 (2H, d, J = 12.0 Hz), 7.63 (1H, d, J = 6.0 Hz), 7.48 (2H, d, J = 6.0 Hz), 7.33 (2H, m), 7.15 (2H, m), 6.66 (1H, d, J = 12.0 Hz), 4.82 (2H, m), 3.46 (2H, t, J = 6.0, 12.0 Hz), 2.22 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.0, 162.5, 160.9, 143.6, 135.1, 135.1, 133.5, 131.0, 130.6, 129.4, 119.7, 115.4, −115.3, 93.5, 70.7, 61.4, 50.6, 31.0. HR-ESI-MS: m/z 350.1346 [M + H]+ calcd for C18H20N2O4, 349.1314.

Synthesis of tert-Butyl (1-(3-((7-Oxo-7H-furo[3,2-g]chromen-4-yl)oxy)propyl)piperidin-4-yl)carbamate (C8)

Light yellow oil, yield 28%. The synthesis method is the same as that of C6, and the raw material is replaced with 4-Boc-aminopiperidine. 1H NMR (400 MHz, CDCl3-d): δ 8.16 (1H, d, J = 12.1 Hz), 7.59 (1H, d, J = 4.0 Hz), 7.15 (1H, s), 6.98 (1H, s), 6.29 (1H, d, J = 12.1 Hz), 4.52 (2H, t, J = 8.0, 12.0 Hz), 3.42 (2H, m), 2.07 (2H, m), 2.02 (2H, m), 1.63 (2H, m), 1.46 (9H, s). 13C NMR (150 MHz, CDCl3-d): δ 161.2, 158.3, 156.1, 152.7, 148.7, 144.9, 139.2, 112.8, 106.7, 105.1, 94.1, 79.6, 70.8, 69.0, 40.1, 37.7, 30.7, 29.4, 28.4 × 3, 21.2. HR-ESI-MS: m/z 443.2109 [M + H]+ calcd for C24H30N2O7, 442.2104.

Synthesis of 4-(3-((4-Methoxybenzyl)amino)propoxy)-7H-furo[3,2-g]chromen-7-one (C9)

Light yellow oil, yield 18%. The synthesis method is the same as that of C7, and the raw material is replaced with 4-methoxybenzylamine. 1H NMR (400 MHz, DMSO-d6): δ 8.28 (1H, d, J = 12.0 Hz), 7.78 (1H, d, J = 4.0 Hz), 7.21 (1H, s), 7.10 (1H, s), 6.28 (1H, d, J = 12.1 Hz), 4.57 (2H, t, J = 8.0, 12.1 Hz), 3.68 (3H, s), 3.31 (4H, m), 3.02 (2H, m), 2.72 (2H, m), 2.45 (1H, m), 2.43–1.76 (8H, m). 13C NMR (150 MHz, DMSO-d6): δ 175.4, 161.8, 158.4, 152.5, 149.0, 145.4, 139.8, 113.5, 111.7, 106.4, 105.0, 93.1, 70.9, 62.1, 54.7, 52.4, 52.2, 50,8, 40.2, 27.4, 26.6. HR-ESI-MS: m/z 380.1412 [M + H]+ calcd for C22H21NO5, 379.1420.

Synthesis of 4-(4-((4-Methoxybenzyl)amino)butoxy)-7H-furo[3,2-g]chromen-7-one (C10)

Light yellow oil, yield 26%. The synthesis method is the same as that of C7. 1H NMR (400 MHz, DMSO-d6): δ 7.63 (2H, d, J = 12.2 Hz), 7.53 (1H, m), 7.39 (2H, d, J = 8.0 Hz), 7.33 (2H, d, J = 8.0 Hz), 7.25 (2H, m), 6.60 (1H, d, J = 12.1 Hz), 4.81 (2H, m), 3.90 (2H, m), 3.48 (2H, s), 3.32 (3H, s) 2.64 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 166.1, 164.6, 162.9, 143.8, 138.2, 131.8, 131.2, 131.3, 131.2, 130.9, 128.6, 118.7, 116.5, 116.1, 106.7, 105.1, 91.4, 50.6, 46.4, 34.3, 31.1. HR-ESI-MS: m/z 364.1484 [M + H]+ calcd for C22H21NO4, 363.1471.

Synthesis of 4-(3-((1-Benzylpiperidin-4-yl)amino)propoxy)-7H-furo[3,2-g]chromen-7-one (C11)

Light yellow oil, yield 31%. The synthesis method is the same as that of C7, and the raw material is replaced with 4-amino-1-benzylpiperidine. 1H NMR (600 MHz, DMSO-d6): δ 8.32 (1H, d, J = 12.0 Hz), 7.72 (6H, d, J = 12.1 Hz), 7.64 (1H, d, J = 6.0 Hz), 7.38 (1H, s), 6.34 (1H, d, J = 12.1 Hz), 4.64 (2H, d, J = 6.0, 12.1 Hz), 3.13 (3H, d, J = 12.1 Hz), 2.96 (2H, s), 2.80 (4H, m), 2.26 (2H, m), 1.76–2.56 (7H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6 158.1, 152.6, 148.9, 146.5, 140.2, 123.9, 122.2, 113.2, 112.8, 106.3, 106.3, 93.8, 70.2, 46.5, 44.3, 26.6, 24.5, 22.9, 22.1. HR-ESI-MS: m/z 433.2059 [M + H]+ calcd for C26H28N2O4, 432.2047.

Synthesis of 4-(3-(4-Aminopiperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (4)

Intermediate 2 (1 g, 3.1 mmol), 4-N-boc-aminopiperidine (2.48 g, 12.38 mmol), K2CO3 (1.7 g, 12.38 mmol), and NaI (232.1 mg, 1.55 mmol) were added in a 100 mL round-bottom flask, heated with acetone (40 mL) as the solvent, and stirred overnight at 60 °C. The reaction solution was purified by silica gel column chromatography, eluted with petroleum ether/ethyl acetate (10:1 to 5:1), the relevant streams were combined, and the solvent was removed under reduced pressure to obtain a light yellow oil product. It was transferred to a 50 mL eggplant-shaped bottle, and 8 mL of ethyl acetate hydrochloride was added and stirred at room temperature for 2 h. Awhite solid was precipitated, filtered, and dried to obtain the final product. Yield 84%, white solid.

Synthesis of (R)-4-(3-(3-Minopiperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (5)

The synthesis method of 5 is the same as that of 4. The raw material is replaced with (S)-3-(BOC amino) piperidine piperidine. Yield 72%, white solid.

Synthesis of (S)-4-(3-(3-Aminopiperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (6)

The synthesis method of 6 is the same as that of 5. The raw material is replaced with (R)-3-(BOC amino) piperidine piperidine. Yield 77%, white solid.

Synthesis of 4-(3-(4-(Benzylamino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (1a)

White solid, yield 57%. 4 (0.2 g, 483.1 μmol) and benzaldehyde (30.7 mg, 365.1 μmol) were taken into a 24 mL round-bottom flask. 5 mL of methanol as a solvent was added, the pH was adjusted to 5–6 with glacial acetic acid, and refluxed at 78 °C for 1 h. After cooling to room temperature, sodium cyanoborohydride (34.4 mg, 547.6 μmol) was added in an ice bath, and the reaction was continued for 4 h. After monitoring by TLC until the reaction was complete, the reaction solution was quenched with saturated aqueous NaHCO3 solution. After the reaction solution was evaporated to dryness under vacuum, it was extracted with dichloromethane and water, and the dichloromethane part was recovered. It was purified by silica gel column chromatography and eluted with dichloromethane and methanol (50:1 to 10:1). 1/1000 triethylamine was added to the mobile phase, and the solvent was removed to obtain a white solid with a yield of 70%. 1H NMR (600 MHz, DMSO-d6): δ 8.10 (1H, d, J = 12.1 Hz), 7.60–7.40 (7H, m), 6.36 (1H, d, J = 12.0 Hz), 4.67 (2H, t, J = 6.0, 12.2 Hz), 3.59 (2H, s), 2.62 (2H, s), 2.39–1.91 (13H, m). 13C NMR (150 MHz, DMSO-d6): δ 161.5, 155.3, 155.4, 132.9, 131.7, 131.2, 130.5, 130.5, 130.2, 129.7, 128.1, 120.3, 120.1, 118.7, 118.6, 116.9, 116.7, 95.4, 78.2, 64.3, 59.5, 58.1, 50.9, 45.9, 43.6, 29.1, 26.7. HR-ESI-MS: m/z 433.2197 [M + H]+ calcd for C26H28N2O4, 433.2122.

Synthesis of 4-(3-(4-((4-Fluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (1b)

Yield 48%, white solid. The synthesis method is the same as that of 1a, and benzaldehyde was replaced with 4-F-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.22 (1H, d, J = 12.1 Hz), 7.82 (3H, m), 7.58 (3H, m) 7.35 (1H, m), 6.35 (1H, d, J = 12.0 Hz), 4.57 (2H,t, J = 6.0, 12.2 Hz), 3.65 (2H, s), 2.75 (2H, m), 2.11–1.78 (9H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 158.0, 152.6, 146.5, 140.0, 135.9, 135.5, 129.7, 129.4, 113.5, 112.8, 112.5, 106.6, 106.1, 1065.9, 93.7, 70.3, 63.1, 60.7, 57.5, 55.5, 52.6, 33.2, 28.6, 25.3. HR-ESI-MS: m/z 451.2081 [M + H]+ calcd for C26H27FN2O4, 451.2028.

Synthesis of 4-(3-(4-((4-Fluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (1c)

Yield 51%, white solid. The synthesis method is the same as that of 1a, and benzaldehyde was replaced with 3-F-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.22 (1H, d, J = 12.0 Hz), 8.03 (1H, m), 7.35–7.34 (3H, m) 7.12 (1H, s), 6.34 (1H, m), 4.64 (2H, t, J = 6.0, 12.0 Hz), 3.67 (2H, s), 1.99–1.23 (13H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.6, 148.7, 146.6, 140.5, 140.1, 130.3, 130.0, 120.6, 117.9, 113.3, 112.9, 112.8, 106.4, 106.2, 94.0, 70.2, 65.5, 55.5, 55.3, 44.0, 42.4, 32.7, 23.2, 21.6. HR-ESI-MS: m/z 451.2084 [M + H]+ calcd for C26H27FN2O4, 451.2028.

Synthesis of 4-(3-(4-((3,4-Difluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (1d)

Yield 42%, white solid. The synthesis method is the same as that of 1a, and benzaldehyde was replaced with 3,4-di-F-benzaldehyde 1H NMR (600 MHz, DMSO-d6): δ 8.26 (1H, d, J = 12.2 Hz), 7.37 (3H, m), 7.13 (3H, m) 6.35 (1H, d, J = 12.1 Hz), 4.62 (2H, t, J = 6.0, 12.1 Hz), 3.53 (2H, s), 3.21 (2H, m), 2.89 (2H, s), 2.22–2.02 (6H, m), 1.61 (2H, m), 2.23–1.18 (3H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.6, 148.8, 146.5, 140.1, 130.6, 130.6, 125.2, 118.7, 116.7, 114.2, 113.2, 112.8, 106.4, 106.1, 93.8 70.0, 61.2, 54.7, 51.3, 46.1, 41.6, 29.5, 28.5, 26.9. HR-ESI-MS: m/z 469.1978 [M + H]+ calcd for C26H26F2N2O4, 469.1934.

Synthesis of 4-(3-(4-((2,4-Difluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (1e)

Yield 37%, white solid. The synthesis method is the same as that of 1a, and benzaldehyde was replaced with 2,4-di-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.34 (1H, d, J = 12.0 Hz), 8.10 (1H, m), 7.86 (1H, m), 7.53 (1H, s), 7.41 (2H, m), 7.32 (1H, m), 6.38 (1H, d, J = 12.0 Hz), 4.94 (2H, m), 3.37–3.55 (4H, m), 2.26 (2H, m), 2.96–2.73 (4H, m). 2.12–1.92 (4H, m). 13C NMR (150 MHz, DMSO-d6): δ 163.5, 161.9, 155.3, 142.4, 130.5, 130.4, 130.1, 125.1, 118.1, 115.6, 115.4, 114.1, 114.0, 110.2, 107.0, 95.0, 70.26, 61.8, 52.6, 51.0, 48.9, 42.7, 32.3, 29.5, 22.6. HR-ESI-MS: m/z 469.1903 [M + H]+ calcd for C26H26F2N2O4, 469.1934.

Synthesis of (S)-4-(3-(3-(Benzylamino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (2a)

Yield 64%, white solid. The synthesis method is the same as that of 1a, and4 was replaced with 5. 1H NMR (600 MHz, DMSO-d6): δ 8.10 (1H, d, J = 12.1 Hz), 7.74 (2H, d, J = 8.0 Hz), 7.42 (1H, s) 7.40 (2H, d, J = 8.0 Hz), 7.37 (1H, m), 7.36 (1H, m), 6.35 (1H, d, J = 12.0 Hz), 4.70 (2H, s), 3.57 (2H, s), 2.57–2.26 (6H, m), 2.15–1.50 (7H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.5, 158.1, 152.6, 148.6, 146.6, 140.1, 140.2, 136.3, 136.0, 136.0, 116.7, 116.4, 113.5, 112.9, 106.4, 106.1, 94.1, 70.0, 57.6, 57.6, 56.5, 56.7, 42.6, 26.3, 22.8, 18.0. HR-ESI-MS: m/z 433.2107 [M + H]+ calcd for C26H28N2O4, 433.2122.

Synthesis of (S)-4-(3-(3-((4-Fluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (2b)

Yield 54%, white solid. The synthesis method is the same as that of 2a, and benzaldehyde was replaced with 4-F-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.29 (1H, d, J = 12.0 Hz), 8.08 (1H, s), 7.40 (1H, s) 7.38 (2H, d, J = 6.0 Hz), 7.19 (1H, m), 7.17 (1H, m), 6.34 (1H, d, J = 12.1 Hz), 4.62 (2H, m), 3.51 (2H, s), 2.35–2.17 (6H, m), 1.56–1.38 (7H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.6, 148.9, 146.5, 141.4, 140.2, 130.6, 125.2, 115.7, 115.6, 114.2, 113.3, 112.8, 106.4, 106.2, 93.8, 70.1, 51.5, 46.0, 46.0, 46.0, 29.5, 26.8, 22.6. HR-ESI-MS: m/z 451.2077 [M + H]+ calcd for C26H27FN2O4, 451.2028.

Synthesis of (S)-4-(3-(3-((3-Fluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (2c)

Yield 51%, white solid. The synthesis method is the same as that of 2a, and benzaldehyde was replacedwith 3-F-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.34 (1H, d, J = 12.2 Hz), 8.11 (1H, d, J = 6.0 Hz), 7.86 (1H, m) 7.53 (1H, s), 7.41 (2H, m), 7.32 (1H, m), 6.38 (1H, d, J = 1.21 Hz), 4.71 (2H, t, J = 6.0, 12.1 Hz), 3.88 (2H, d, J = 6.0 Hz), 3.25 (2H, m), 2.36–1.17 (5H, m), 1.94–1.43 (6H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.6, 148.8, 146.6, 140.1, 140.0, 136.3, 126.0, 116.6, 116.4, 113.5, 113.5, 112.9, 106.5, 106.1, 94.2, 70.0, 66.8, 57.6, 56.7, 53.2, 42.5, 26.1, 22.6, 18.0. HR-ESI-MS: m/z 451.2093 [M + H]+ calcd for C26H27FN2O4, 451.2028.

Synthesis of (S)-4-(3-(3-((3,4-Difluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (2d)

Yield 49%, white solid. The synthesis method is the same as that of 2a, and benzaldehyde was replaced with 3,4-di-F-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 7.68 (1H, s), 7.38 (2H, s), 7.21–6.98 (4H, s), 6.45 (1H, d, J = 12.1 Hz), 4.53 (2H, t, J = 6.0, 12.1 Hz), 3.50 (2H, s), 3.21–3.03 (2H, m), 2.63 (1H, m), 2.41–1.52 (12H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.5, 158.1, 152.6, 150.5, 146.7, 140.3, 125.2, 125.0, 125.0, 123.1, 118.8, 118.7, 113.6, 112.9, 106.6, 106.1, 94.2, 70.0, 61.7, 57.6, 56.6, 53.6, 42.6, 28.6, 26.0, 22.6. HR-ESI-MS: m/z 469.1979 [M + H]+ calcd for C26H26F2N2O4, 469.1934.

Synthesis of (S)-4-(3-(3-((2,4-Difluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (2e)

Yield 32%, white solid. The synthesis method is the same as that of 2a, and benzaldehyde was replaced with 2,4-di-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 7.82 (1H, d, J = 12.0 Hz), 7.35 (1H, d, J = 12.0 Hz), 7.10 (4H, m), 6.76 (1H, d, J = 12.2 Hz), 4.32 (2H, t, J = 6.0, 12.2 Hz), 3.44 (2H, s), 3.33–3.21 (2H, m), 2.73 (2H, m), 2.34 (2H, m), 1.96 (3H, m), 1.67 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.5, 158.1, 152.6, 148.8, 146.7, 140.6, 140.3, 140.1, 137.8, 137.7, 113.4, 112.9, 111.7, 106.5, 106.1, 105.5, 94.1, 70.0, 60.1, 57.5, 56.3, 54.2, 42.5, 25.8, 22.8, 18.1. HR-ESI-MS: m/z 469.1968 [M + H]+ calcd for C26H26F2N2O4, 469.1934.

Synthesis of (R)-4-(3-(3-(Benzylamino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (3a)

Yield 43%, white solid. The synthesis method is the same as that of 1a, and 5 was replaced with 7. 1H NMR (600 MHz, DMSO-d6): δ 8.22 (1H, d, J = 12.0 Hz), 8.09 (1H, d, J = 1.21 Hz), 7.52–7.33 (5H, m), 6.38 (1H, d, J = 12.0 Hz), 4.71 (2H, s), 3.82 (2H, s), 2.09–1.91 (4H, m), 1.42–1.23 (7H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.5, 158.1, 152.6, 148.9, 146.5, 141.4, 140.2, 130.6, 125.2, 115.7, 115.6, 114.2, 113.3, 112.8, 106.4, 106.2, 93.8, 70.2, 61.5, 54.6, 52.8, 52.7, 42.8, 29.5, 26.8, 22.6. HR-ESI-MS: m/z 433.2173 [M + H]+ calcd for C26H28N2O4, 433.2122.

Synthesis of (R)-4-(3-(3-((3-Fluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (3b)

Yield 45%, white solid. The synthesis method is the same as that of 3a, and benzaldehyde was replaced with 4-F-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.27 (1H, d, J = 8.0 Hz), 8.01 (1H, m), 7.39 (1H, s) 7.37 (1H, m), 7.17(3H, m), 7.08 (1H, m), 6.35 (1H, d, J = 8.0 Hz), 4.61 (2H, t, J = 6.0, 12.1 Hz), 3.52 (2H, m), 2.82 (2H, m), 2.29–2.17 (5H, m), 1.94–1.43 (6H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.5, 158.1, 152.6, 148.9, 146.5, 141.4, 140.2, 130.6, 125.2, 115.7, 115.6, 114.2, 113.3, 112.8, 106.4, 106.2, 93.8, 70.2, 61.5, 54.6, 52.8, 52.7, 42.8, 29.5, 26.8, 22.6. HR-ESI-MS: m/z 451.2091 [M + H]+ calcd for C26H27FN2O4, 451.2028.

Synthesis of (R)-4-(3-(3-((4-Fluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (3c)

Yield 35%, white solid. The synthesis method is the same as that of 3a, and benzaldehyde was replaced with 3-F-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.10 (1H, d, J = 12.1 Hz), 7.74 (2H, d, J = 8.0 Hz), 7.42 (1H, s) 7.40 (2H, d, J = 8.0 Hz), 7.37 (1H, m), 7.36 (1H, m), 6.35 (1H, d, J = 12.1 Hz), 4.70 (2H, s), 3.57 (2H, s), 2.57–2.26 (6H, m), 2.15–1.50 (7H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.5, 158.1, 152.6, 148.6, 146.6, 140.1, 140.2, 136.3, 136.0, 136.0, 116.7, 116.4, 113.5, 112.9, 106.4, 106.1, 94.1, 70.0, 57.6, 57.6, 56.5, 56.7, 42.6, 26.3, 22.8, 18.0. HR-ESI-MS: m/z 451.2067 [M + H]+ calcd for C26H27FN2O4, 451.2028.

Synthesis of (R)-4-(3-(3-((3,4-Difluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (3d)

Yield 29%, white solid. The synthesis method is the same as that of 3a, and benzaldehyde was replaced with 3,4-di-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.10 (1H, d, J = 12.1 Hz), 7.84 (2H, m), 7.63 (2H, s) 7.56 (2H, m), 7.42 (2H, m), 6.35 (1H, d, J = 12.0 Hz), 4.70 (2H, s), 3.07–2.80 (6H, m), 2.32–1.80 (10H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.5, 158.1, 152.6, 150.5, 146.7, 140.3, 125.2, 125.0, 125.0, 123.1, 118.8, 118.7, 113.6, 112.9, 106.6, 106.1, 94.2, 70.0, 61.7, 57.6, 56.6, 53.6, 42.6, 28.6, 26.0, 22.6. HR-ESI-MS: m/z 469.1962 [M + H]+ calcd for C26H26F2N2O4, 469.1934.

Synthesis of (R)-4-(3-(3-((2,4-Difluorobenzyl)amino)piperidin-1-yl)propoxy)-7H-furo[3,2-g]chromen-7-one (3e)

Yield 25%, white solid. The synthesis method is the same as that of 3a, and benzaldehyde was replaced with 2, 4-di-benzaldehyde. 1H NMR (600 MHz, DMSO-d6): δ 8.08 (1H, d, J = 12.0 Hz), 7.39 (1H, d, J = 6.0 Hz), 7.37 (2H, s) 7.17 (2H, d, J = 8.0 Hz), 7.08 (1H, m), 6.35 (1H, d, J = 12.1 Hz), 4.61 (2H, t, J = 6.0, 12.0 Hz), 3.51 (2H, s), 2.29–2.17 (6H, m), 1.94–1.48 (6H, m). 13C NMR (160 MHz, DMSO-d6): δ 160.6, 158.1, 152.6, 148.8, 146.7, 140.3, 140.2, 137.8,137.7, 137.6, 113.4, 113.2, 112.9, 111.5, 106.5, 106.1, 105.6, 94.2, 70.0, 61.7, 57.3, 56.5, 56.4, 54.3, 42.6, 25.8, 22.7, 18.1. HR-ESI-MS: m/z 469.1972 [M + H]+ calcd for C26H26F2N2O4, 469.1934.

Synthesis of 4-(3-((1-Benzylpiperidin-4-yl)amino)propoxy)-7H-furo[3,2-g]chromen-7-one (4a)

Yield 39%, white solid. 4-Boc-aminopiperidine (2 g, 10 mmol), benzyl bromide (3.6 mL, 30 mmol), K2CO3 (12.4 g, 90 mmol), NaI (0.75 g, 5 mmol), and acetone (40 mL) were added to a 100 mL round-bottom flask, refluxed, and stirred overnight. The reaction solution was purified by silica gel column chromatography, eluted with petroleum ether/ethyl acetate (10:1 to 5:1), and the solvent was removed under reduced pressure to obtain a light yellow oily product which was transferred to a 50 mL eggplant flask, 10 mL of ethyl acetate hydrochloride was added, stirred at room temperature for 2 h, and a white solid was precipitated which was filtered and dried to obtain intermediate 5. Then, 3 (100 mg, 0.3 mmol), 5 (610 mg, 2.7 mmol), K2CO3 (372 mg, 2.7 mmol), and NaI (0.15 g, 0.1 mmol) were taken, acetone (40 mL) was used as a solvent, heated at 60 °C, and stirred overnight, and the reaction solution was purified by silica gel column chromatography to obtain 4a. 1H NMR (600 MHz, DMSO-d6): δ 8.32 (1H, d, J = 12.1 Hz), 7.79 to 7.72 (6H, m), 7.44 (1H,d, 6 Hz) 7.38 (1H, s), 6.34 (1H, d, J = 12.1 Hz), 4.64 (2H, t, J = 6.0, 12.1 Hz), 3.13 (2H, s), 2.98 (2H, s), 2.80 (4H, m), 2.27 (2H, m), 1.76–1.56 (4H, m). 13C NMR (150 MHz, DMSO-d6): δ 164.5, 156.3, 155.3, 134.3, 133.8, 130.8, 130.8, 129.6, 128.1, 127.6, 109.7, 106.1, 92.6, 71.4, 66.0, 61.7, 56.5, 55.7, 51.7, 34.2, 27.6, 19.4. HR-ESI-MS: m/z 433.2143 [M + H]+ calcd for C26H28N2O4, 433.2122.

Synthesis of 4-(3-((1-(3-Fluorobenzyl)piperidin-4-yl)amino)propoxy)-7H-furo[3,2-g]chromen-7-one (4b)

Yield 27%, white solid. The synthesis method is the same as that of 4a, and benzyl bromide was replaced with 4-fluorobenzyl bromide. 1H NMR (600 MHz, DMSO-d6): δ 8.08 (1H, d, J = 12.1 Hz), 7.58–7.39 (7H, m), 7.42 (1H, s), 6.34 (1H, d, J = 12.1 Hz), 4.61 (2H, t, J = 6.0, 12.1 Hz), 3.81–3.39 (9H, m), 2.48–2.18 (6H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.5, 148.6, 146.6, 140.5, 140.1, 135.4, 124.3, 124.3, 116.6, 116.3, 113.3, 112.8, 106.4, 106.1, 94.0, 70.1, 64.2, 60.2, 55.4, 55.0, 44.1, 36.3, 23.2, 22.8. HR-ESI-MS: m/z 451.2041 [M + H]+ calcd for C26H27FN2O4, 451.2028.

Synthesis of 4-(3-((1-(4-Fluorobenzyl)piperidin-4-yl)amino)propoxy)-7H-furo[3,2-g]chromen-7-one (4c)

Yield 22%, white solid. The synthesis method is the same as that of 4a, and benzyl bromide was replaced with 3-Fluorobenzyl bromide. 1H NMR (600 MHz, DMSO-d6): δ 8.29 (1H, d, J = 12.0 Hz), 8.08 (2H, d, J = 6.0 Hz), 7.40 (1H, s) 7.39 (2H, d, J = 6.0 Hz), 7.18 (2H, m), 7.17 (1H, m), 6.34 (1H, d, J = 12.0 Hz), 4.62 (2H, t, J = 6.0, 12.0 Hz), 3.51 (2H, s), 2.35–1.49 (13H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 152.5, 148.7, 146.6, 140.5, 140.1, 130.3, 130.0, 120.6, 117.9, 113.3, 112.9, 112.8, 106.4, 106.1, 94.0, 70.2, 65.5, 55.7, 55.3, 44.0, 42.4, 32.7, 23.1, 21.6. HR-ESI-MS: m/z 451.2068 [M + H]+ calcd for C26H27FN2O4, 451.2028.

Synthesis of 4-(3-((1-(3,4-Difluorobenzyl)piperidin-4-yl)amino)propoxy)-7H-furo[3,2-g]chromen-7-one (4d)

Yield 19%, white solid. The synthesis method is the same as that of 4a, and benzyl bromide was replaced with 3, 4-fluorobenzyl bromide. 1H NMR (600 MHz, DMSO-d6): δ 8.10 (1H, s), 8.78 (2H, d, J = 6.0 Hz), 7.62 (2H, m) 7.44 (2H, m), 6.33 (1H, d, J = 12.1 Hz), 4.71 (2H, s), 378 (2H, s), 2.66–1.99 (13H, m). 13C NMR (150 MHz, DMSO-d6): δ 159.5, 157.0, 151.5, 147.6, 145.6, 139.3, 139.0, 126.9, 124.3, 121.9, 112.4, 111.8, 105.4, 105.1, 105.0, 93.1, 69.2, 59.2, 54.3, 53.7, 51.6, 42.8, 28.4, 22.1, 21.6. HR-ESI-MS: m/z 469.2029 [M + H]+ calcd for C26H26F2N2O4, 469.1934.

Synthesis of 4-(3-((1-(2,4-Difluorobenzyl)piperidin-4-yl)amino)propoxy)-7H-furo[3,2-g]chromen-7-one (4e)

Yield 15%, white solid. The synthesis method is the same as that of 4a, and benzyl bromide was replaced with 32, 4-Fluorobenzyl bromide. 1H NMR (600 MHz, DMSO-d6): δ 8.10 (1H, s), 7.80 (1H, m), 7.51 (1H, m) 7.45–7.36 (4H, m), 6.36 (1H, d, J = 12.0 Hz), 4.61 (2H, t, J = 6.0, 12.1 Hz), 3.84 (2H, s), 2.47–2.05 (13H, m). 13C NMR (150 MHz, DMSO-d6): δ 160.6, 158.1, 158.1, 148.7, 146.7, 146.7, 140.4, 137.6, 113.4, 113.4, 113.4, 112.9, 112.9, 106.5, 106.4, 106.2, 94.1, 70.2, 60.2, 55.3, 54.9, 52.9, 43.8, 23.3, 22.87, 22.5. HR-ESI-MS: m/z 469.1948 [M + H]+ calcd for C26H26F2N2O4, 469.1934.

Biological Assay

Inhibition of BACE1

The fluorescence resonance energy transfer method was used to study the inhibition of BACE1 activity of notopterol derivatives as our previous study.[25] LY2811376, a BACE1 inhibitor, was purchased from MedChem Express (Shanghai, China) as a positive drug. Recombinant human BACE-1 (rhBACE1) was purchase from Sino biological (Beijing, China), and its fluorogenic peptide substrate (Mca-SEVNLDAEFRK(Dnp)RR-NH2) was purchased from ChinaPeptides (Suzhou, China). In brief, rhBACE1 was diluted with assay buffer (0.1 M sodium acetate, pH 4.0), then mixed with heparin solution (4 ng/uL) in equal volume in a black 384-well plate, and incubated at 37 °C for 30 min. The fluorescent polypeptide substrate was diluted to 20 mM by assay buffer. 15 μL of rhBACE1–heparin was mixed into the 384-well plate, 7.5 μL of substrate (20 μM) and 7.5 μL of notopterol derivative (80 μM) with different concentrations were added, and the reaction was carried out for 60 min at rt. The fluorescence value was measured by Fluoroskan (Thermo, USA) at an excitation wavelength of 320 nm and emission wavelength of 405 nm. The fluorescence intensity of the blank and the positive drug were recorded, and the blank background signal was subtracted to calculate the percentage inhibition of the compound.

Inhibition of GSK3β

Recombinant human GSK3β (rh GSK3β) was purchased from Sino biological (Beijing, China) and its prephosphorylated polypeptide substrate GSM (YRRAAVPPSPSLSRHSSPHQ-pS-EDEEE) was purchased from Synpeptide (Nanjing, China). Kinase-Glo system (Promega, USA) was used to determine the remaining ATP. Referring to the instructions of ADP-Glo kinase assay, the enzyme, substrate, ATP, and inhibitors were diluted in kinase buffer in the black 384-well plate. The ATP was mixed with the substrate in equal volume (substrate: 0.5 μg/μL, ATP: 5 mM). 2 μL of enzyme (25 ng/μL), 1 μL of notoperol derivatives (5% DMSO), and 2 μL of substrate/ATP were added to a white 384-well plate. After incubating at room temperature for 60 min, 5 μL of ADP-Glo reagent was added and incubated at rt for 40 min. Then, 10 μL of kinase detection reagent was added. Luminescence (integration time 0.5–1 s) was recorded after 30 min. Tideglusib was obtained from Bidepharm (Shanghai, China) as the positive drug.

AChE Inhibitory Activity

For the acetylcholinesterase inhibition experiment, the previous literature was referred,[29] using the modified Ellman method to test the AChE inhibitory activity of the compound. Electric eel-derived AChE, thioacetylcholine iodide (ATChI), and 5,5-dithiobis(2-nitrobenzoic acid) (DTNB) were purchased from Sigma-Aldrich, USA. Test compounds were dissolved in buffer solution (50 mM Tris–HCl, pH = 8.0, 0.1 M NaCl, 0.02 M MgCl2·6H2O) containing 1% DMSO. In the wells of a 96-well plate, 160 μL of 1.5 mM DTNB and 50 μL of AChE (0.22 U/mL prepared in 50 mM Tris-HCl, pH = 8.0, 0.1% w/v fetal bovine serum, BSA) were added, respectively. 10 μL of compound solution was added and incubated at 37 °C for 6 min. Then, 10 μL of DTNB (2 mM) and 10 μL of ATChI (15 mM) were added, and the absorbance was read at 405 nm after 20 min at 37 °C. In the blank wells, buffer solution was used to replace the drug, and tacrine was used as a positive control at a concentration of 1.0 μM. tThe inhibition rate of the sample to be tested was calculated according to the following formula

PAMPA-BBB Assay

The method of PAMPA-BBB assay was carried out according to the method of our previous study.[23] Ten commercial drugs were used to validate the protocol and purchased from Solarbio Life Sciences. Dodecane was obtained from Sigma-Aldrich. The porcine brain lipid (PBL) was purchased from Avanti Polar Lipids. The donor 96-well filter microplate with a PVDF membrane (pore size 0.45 μM) and acceptor indented 96-well microplate were purchased from Millipore. Commercial drugs and test compounds were initially dissolved in DMSO at a concentration of 5 mg/mL. Subsequently, they were diluted 200-fold with a solution of PBS (pH 7.4 ± 0.1)/EtOH (70/30, v/v) to give a final concentration of 100 μg/mL. The filter membrane of the donor microplate was coated with 4 μL of PBL in dodecane (5 mg/mL). Then, 200 μL of diluted compound solution was added into the donor wells and 300 μL of PBS/EtOH (70/30, v/v). The donor filter plate was carefully placed on the top of the acceptor plate to form a “sandwich” assembly to make the membrane contact with buffer solution. The sandwich was put undisturbed at 25 °C. After incubation for 20 h, the donor plate was carefully removed; the concentrations of test compounds in the donor and acceptor wells were measured with a UV plate spectroscopy reader.

Docking and MD Simulations

The operation and parameter setting of docking were referred to our previous study.[27] The X-ray crystal structure of BACE1 (5CLM), GSK3β (4PTC), and AChE (4EY7) crystallographic structures was downloaded from RCSB protein data bank (PDB). Prior to docking, the downloaded protein file was prepared by Schrödinger’s Protein Preparation Wizard. Afterward, the OPLS_2005 force field was used to optimize theprotein energy and eliminate steric hindrance.[28] Finally, the SP Glide method was used to dock the molecules in the prepared data set to the active site cavity of the BACE1 and GSK3β proteins to obtain the interaction model.

MD simulations were performed as described in previous studies.[29,30] The OPLS_2005 force field[31] was used to minimize the energy of complex systems with a maximum interaction setting of 2000 and a convergence threshold of 1.0 kcal/mol/Å. Before starting the simulation, the system performed a 10 ns NPT simulation at a temperature of 300 K set by the nose-Hulf thermostat and a pressure of 1.01325 bar set by the Martyna–Tobias–Klein constant pressure device to relax the composite.[32] Under the NPT system, MD simulations were run for 100 ns, energy and trajectory were recorded every 1.2 and 4.8 ps, respectively, and the resulting data were used for statistical analysis. Potential energy (U), rmsd, root mean square fluctuation, and ligand–protein interactions were monitored to determine docking complex stability.

Animal Treatment

48 male mice, 7–8 weeks old, weighing 22–26 g, were purchased from HFK Bio (Beijing, China) and were randomly divided into 6 groups. The animals were fed under standard conditions of a 12 h light/dark cycle at 22–24 °C with free access to food and water. 1c was dissolved in 5% DMSO, the suspension was mixed with 0.5% CMC-Na, and the corresponding concentrations were prepared for intragastric administration. Donepezil was purchased from Bide Pharmaceuticals (Shanghai, China) as the positive drug, and the gavage dose was 5 mg/kg. After injection of Aβ42 into the lateral ventricle, mice started to gavage 1c at 2.5, 5, and 10 mg/kg for 7 consecutive days, followed by Morris water maze assay (also administered during the period), and donepezil was administered in parallel as a positive drug. In addition, a group of control groups was set up and given the same vehicle by gavage.

Brain Stereotactic Injection

The mice were anesthetized by intraperitoneal injection of sodium pentobarbital (7.5 mg/mL), and then the top of the mouse head was shaved, fixed on the brain stereotaxic apparatus, and the incisor rod was adjusted so that the top was horizontal. The skin on the top of the mouse head was disinfected and then incised along the midline, the fascia was peeled off, and the surface of the skull was rubbed with a sterile cotton swab and a small amount of hydrogen peroxide to expose the Bregrna (anterior fontanelle) point. Lateral ventricle positioning coordinates were selected as mediolateral = ±1.1 mm, anteroposterior = −0.5 mm, and dorsoventral = −3.0 mm (the depth of needle insertion takes the dura as the starting point). After the injection needle reaches the position, a microinjection pump was used to inject 3 μL of Aβ42 (2.5 μg/μL) at a speed of 3 μL/10 min. After injection, the needle was pulled out after staying for 3 min. The skin incision was sutured, intramuscular injection of penicillin was used to prevent infection, and the mice were returned to the cage for feeding recovery.

Blood Biochemical Assay

The detection methods of ALT, AST, and BUN were referred to our previous study.[33] Mice were anesthetized by intraperitoneal injection of chloral hydrate after fasting for 8 h. Blood which was obtained from the eyeball was separated and centrifuged at 3000 rpm for 10 min. Then, we collected the supernatant and stored at −80 °C. ALT, AST, and BUN levels were determined by the standard operating procedures of the Beckman Coulter Biochemical Analyzer (AU5800).

HE Staining

Methods of tissue sample processing are described in our previous study.[34] Briefly, after the mice were sacrificed, the liver and kidney were separated and fixed with 4% paraformaldehyde. Then, they were embedded to prepare paraffin samples. The wax was sliced to a 3 μm thin section by Leica RM2235. The slices were dewaxed at 60 °C for 1 h and soaked in gradient alcohol. The sections were stained with HE and sealed with a neutral resin. The images were taken on a Nikon microscope (Eclipse Ni-E, Japan).

PK Study

The drug configuration was described as described previously. 1c (100 mg/kg) was orally administered to SD rat or tail vein injection (10 mg/kg). Eyeball blood was collected at 0, 0.25, 0.5, 1, 2, 4, 6, 8, and 12 h after administration of 1c (n = 5 for each group). The intrinsic standard (terfenadine) was added to blood or tissue homogenate samples, and 1c was extracted by 3-fold acetonitrile quenching. The plasma was isolated from the blood samples by centrifugation at 12,000 rpm for 15 min at 4 °C. The supernatant was directly injected into the LC/MS/MS system (MD3200, Applied Biosystems). The reversed phase column and gradient elution of water/1% formic acid and acetonitrile/1% formic acid were used. 1c was measured by positive electrospray ionization and MRM quantification. Das 3.0 software was used for analyzing data. See the Supporting Information for details.

Morris Water Maze Assay

The Morris water maze experiment is divided into two parts: directional navigation and space exploration. The pool was divided into four quadrants, the mice were placed facing the pool wall, put into the water from the first and third quadrant entry points, respectively, and the time from entering the water to finding a platform hidden 1 cm underwater and staying was recorded, that is, the incubation period. If the mouse did not find the platform within 60 s, the experimenter led it to the platform, and its latency was recorded as 60 s. Training was done once a day for a total of 6 days. According to the swimming trajectory of the mouse in the water to search for the platform, its search strategy is determined each time, so as to judge its learning ability. On the seventh day of the experiment, the platform was removed, and the space exploration experiment was carried out. The water entry point of the platform was selected relative to the quadrant (second quadrant), the mouse was put into the water facing the pool wall, the swimming trajectory of the mouse was recorded in 60 s, and the number of times the mouse crosses the platform was recorded to evaluate the space exploration ability of the mouse.

Western Blot

Tissue samples were processed according to methods previously reported.[25] Briefly, cortical and hippocampal tissues of 3 mice in each group were added to 10-fold volume of RIPA buffer (Wanlei Bio, China) containing protease cocktail (Sigma-Aldrich, USA). They were homogenized with a homogenizer (60 Hz, 1 min) and incubated on ice for 30 min to completely lyse the cells. The homogenate was then centrifuged at 20,000g min–1 for 30 min at 4 °C. The supernatant was collected, and the protein concentration was determined using the BCA protein assay kit. 4× loading buffer was added and mixed and then boiled at 100 °C for 10 min. The target protein samples were separated by 12% SDS polyacrylamide gel electrophoresis, and then the separated proteins were transferred to the PVDF membrane by an electrotransfer device. The transferred PVDF membrane was blocked with 5% nonfat milk powder at room temperature for 2 h and incubated with ADAM17 (1000:1) and BACE1 (1000:1) antibodies overnight at 4 °C. The PVDF membrane was washed with TBST for 5 min × 3 times, the secondary antibody was added and incubated at room temperature for 1 h, and the membrane was washed with TBST for 5 min × 3 times. Then, ECL chemiluminescent solution was used to develop, take the image with Bio-Rad gel imaging system, and analyze the gray value of the band with ImageJ software.