Study on the Synthesis and Biological Activities of N-Alkylated Deoxynojirimycin Derivatives with a Terminal Tertiary Amine

A series of N-alkylated deoxynojirimycin (DNJ) derivatives connected to a terminal tertiary amine at the alkyl chains of various lengths were prepared. These novel synthetic compounds were assessed for preliminary glucosidase inhibition and anticancer activities in vitro. Potent and selective inhibition was observed among them. Compound 7d (IC50 = 0.052 mM) showed improved and selective inhibitory activity against β-glucosidase compared to DNJ (IC50 = 0.65 mM). In addition, analysis of the kinetics of enzyme inhibition by using Lineweaver–Burk plots indicated that 7d inhibited β-glucosidase in a competitive manner, suggesting that 7d was expected to bind to the active site of β-glucosidase. Compounds 8b and 8c were found to be moderate and selective inhibitors of α-glucosidase. Nevertheless, none of compounds inhibited the growth of B16F10 melanoma cells.


Introduction
Glucosidases are enzymes which catalyze the hydrolysis of glycosidic bonds in oligosaccharides or glycoconjugates, playing a vital role in the digestion of carbohydrates and in the processing of glycoproteins and glycolipids. 1 Glucosidases are also involved in carbohydrate-mediated diseases such as diabetes, 2 tumor metastasis, 3 viral infections, 4 and lysosomal storage diseases. 5 Inhibitors of α-glucosidase can significantly decrease postprandial blood glucose levels 6 and promote glycoprotein misfolding in the endoplasmic reticulum (ER). 7 In mammals, β-glucosidase enables hydrolysis of glucosylceramide into ceramide and glucose, which is in part performed by β-glucocerebrosidases (GBA1 or GCase) 8 and GBA2. 9 Gaucher disease, the most common lysosomal storage disease, is caused by mutations in the β-glucocerebrosidase (GBA1) gene. Inhibitors of β-glucosidase could reduce the biosynthesis of glycolipids to balance the deficient activity of β-Gcase. 10 In tumor cells, oligosaccharides on the surface of tumor cells play an important role in expression of the malignant phenotype and the metastatic spread of tumor cells. The synthesis of these oligosaccharides in endoplasmic reticulum and Golgi is dependent on carbohydrate processing enzymes such as glycosidases. Therefore, specific glycosidase inhibitors may be candidates for cancer chemotherapy. 11,12 Among the families of glycosidase inhibitors reported so far, iminosugars are particularly notable. They are carbohydrate mimetics where the endocyclic oxygen has been replaced by a nitrogen atom. [13][14][15] Their structures can mimic transition-state analogues of glycosidases, which interact with two carboxylic acid units to form strong ions and catalyze the cleavage of the glycoside bonds. 1 Their most famous representative is the naturally occurring 1-deoxynojirimycin 1. 2 Some N-alkylated DNJ derivatives, like N-hydroxyethyl-DNJ 16 2 (miglitol, an intestinal α-glucosidase inhibitor), and N-butyl-DNJ 17 3 (miglustat, a glucosylceramide synthase inhibitor) have been approved for the treatment of diabetes-type 2 and Gaucher disease, respectively. Compound 6 3 not only inhibited α-glucosidase (Bacillus stearothermophilus), BAEC growth and migration, but also suppressed the growth of A549 cells ( Figure  1). Nevertheless, despite extensive synthesis and investigations of highly bioactive iminosugars, a remaining drawback is their limited selectivity on glucosidases, and this leads to some side effects when applied therapeutically. For example, N-butyl-DNJ 3 ( Figure 1) can inhibit some other enzymes nonrelated to lysosomal storage disease, such as sucrase, maltase, α-glucosidase I and II. 18 Obviously, improving the selectivity of iminosugars as glycosidase inhibitors is a challenging goal.
Modification or variation of a known iminosugar inhibitor, especially a natural product, is a feasible strategy to obtain more selective and stronger inhibitors. Generally, there are two main strategies for modification of iminosugars: introduction of different alkyl groups on the amino group and alterations of the ring hydroxyl residues. 19 It has been demonstrated that the potency of DNJ derivatives could be increased by introducing a hydrophobic group on the nitrogen atom of DNJ using a heteroatom linker and a carbon chain spacer. Moreover, lengthening of the alkyl chain and an increase in the size of the hydrophobic group would be also beneficial for the glucosidase inhibition. These types of modifications can be seen in the design of compounds 4, 20 5 21 and 6 ( Figure 1). 3 Our group had done some work on the modification of DNJ, such as the synthesis of C-6 deutero DNJ, a potent α-glucosidase and the optimization of DNJ synthetic route. 22,23 And as a part of our ongoing program devoted to the development of new glucosidase inhibitors, we embarked on a strategy starting from DNJ as the lead compound. The key DNJ scaffold was connected to a terminal tertiary amine through introduction of alkyl chains of various length. And the introduction of a nitrogen atom may lead to a polarization different from that of oxygen atom. 24,25 The work reported herein describes the synthesis and biological evaluation of a small library of DNJ derivatives in which the length of the alkyl chain and the size and nature of the terminal tertiary amine substituents have been studied.

1. Materials and Methods
All reagents and solvents were purchased from commercial suppliers and used without further purification. Reactions progression was monitored by Thin Layer Chromatography (TLC) using silica gel GF 254 plates (0.2 mm thickness), spots were detected under UV-light (λ = 254 nm). Visualization of the deprotected iminosugar was accomplished by exposure to iodine vapour. Flash column chromatography was carried out by silica gel (200-300 mesh). NMR spectra were recorded on Bruker Avance III 500 MHz spectrometer using CDCl 3 or D 2 O as solvents. Chemical shifts are reported in ppm. High resolution mass spectra (HRMS) were recorded by direct injection on a mass spectrometer (Thermo Scientific LTQ Orbitrap XL) equipped with an electrospray ion source in positive mode. The following abbreviations have been used to describe the signal multiplicity: br (broad), s (singlet), d (doublet), t (triplet), q (quartet), h (hextet), m (multiplet), dd (doublet of doublets), dt (doublet of triplets).

General Procedure A
Nucleophilic substitution on a nitrogen atom. The starting material (1 mM) was mixed with N-bromophthalimide (2 mM) and K 2 CO 3 (3 mM) in DMF (10 mL). The mixture was heated at 100 °C for 24 h. After cooling, the mixture was poured into water and extracted into ethyl acetate. The organic layer was dried over Na 2 SO 4 and concentrated. The residue was purified by flash column chromatography (10:1→3:1; PE:ethyl acetate).
General Procedure E Catalytic hydrogenolysis. To a solution of the benzylated intermediate (1 mmol) in EtOH was added Pd (10%)/C (100 mg) and the mixture stirred under an atmosphere of hydrogen at room temperature for 24 h. The catalyst was filtered off, the solvents removed under reduced pressure and the residue purified by flash column chromatography (8:2:0.1→6:4:0.1; n-propanol:H 2 O:NH 4 OH).

Glucosidase Inhibitory Assays
α-Glucosidase (yeast), β-glucosidase (sweet almonds), and α-mannosidase (jack bean) was purchased from Sigma. 1-Deoxynojirimycin, para-nitrophenyl α-D-glucopyranoside, para-nitrophenyl β-D-glucopyranoside and para-nitrophenyl α-D-mannosidase were also purchased from Sigma. Inhibitory potencies were carried out by spectrophotometrically measuring the residual hydrolytic activities of the glycosidases on the corresponding para-nitrophenyl glycoside substrates. The α-glucosidase, 26 β-glucosidase assays 27 were performed in 50 mM phosphate buffer, pH 6.8 at 37 °C. The α-mannosidase assay 28 was performed in 50 mM citrate buffer, pH 5.5 at 37 °C. The test compounds were pre-incubated with the enzyme solutions and buffered in a disposable 96-well microtiter plate at 37 °C for 15 min. Next, the reactions were initiated by the addition of 20 µL of a solution of the corresponding para-nitrophenyl glycoside substrates. After the reaction mixture was incubated at 37 °C for 15 min. Thereupon, it was quenched by adding 80 µL Na 2 CO 3 (0.2 mol/L). Enzymatic activity was quantified by measuring the absorbance at 405 nm using a BioTek µQuant Microplate Spectrophotometer. Each experiment was performed in triplicate. IC 50 values were determined graphically with GraphPad Prism (version 8.0).

3. Kinetics of Enzyme Inhibition
Inhibition constant (K i ) measurement was performed in 50 mM phosphate buffer (pH 6.8) at 37 °C, using para-nitrophenyl β-D-glucopyranoside as the substrate. The assay was initiated by adding β-glucosidase (K m = 3.5 mM) to a solution of the substrate (concentrations used: 0.875 mM, 1.75 mM, 3.5 mM, 7 mM, 10.5 mM) in the presence of inhibitors (concentrations used: 0 mM, 0.1 mM, 0.2 mM). After the reaction mixture was incubated at 37 °C for 15 min, it was quenched by adding 80 µL Na 2 CO 3 (0.2 mol/L). The absorbance of 4-nitrophenol released from the substrate was read at 405 nm.

4. Cell Culture and Inhibition of Proliferation B16F10 Cells 29
The mouse B16F10 melanoma cell line, which is derived from C57BL/6 mice was purchased from KeyGen Biotech (Nanjing, China). The cell line was cultured in DMEM supplemented with fetal bovine serum (10%), penicillin (100 U/mL) and streptomycin (100 µg/mL) at 37 °C in humidified 5% CO 2 atmosphere. Media was replenished every third day. B16F10 cells were seeded on 96well microtiter plates in DMEM supplemented with 10% FBS and incubated overnight. The compounds (1 mM, 0.05 mM) were then added to the cells and cultured for another 48 h. Each treatment was performed in six well replicates. MTT reagent (Sigma Aldrich) was added to each well incubated for 4 h at 37 °C. After the cell culture medium was removed, formazan crystals in adherent cells were dissolved in 200 µL DMSO and the absorbance of the formazan solution was measured at 570 nm.

1. Chemistry
The target compounds were prepared from the key intermediate 11 through reductive amination or double nucleophilic substitution, respectively ( Figure 2). The synthesis of compound 11 commenced from 2,3,4,6-tet-ra-O-benzyl-1-deoxynojirimycin 9 which was prepared according to previously published procedures in four steps. 28 Treatment of O-benzyl protected DNJ 9 with N- (4-bromobutyl)phthalimide or N- (4-bromoethyl) phthalimide in the presence of K 2 CO 3 in DMF afforded N-phthalyl protected DNJ 10 (Scheme 1). The intermediate 10 was then converted into primary amide 11 by a hydrazinolysis reaction using N 2 H 4 in EtOH.
A generalized synthetic approach to the derivatives 7 and 8 was shown in Scheme 2. The reductive amination of 11 with HCHO-HCOOH gave compounds 12a and 13a. For compounds 12 and 13 which beared 5-and 6-membered rings, double nucleophilic substitution reaction was performed on primary amine 11 in basic conditions. All the intermediates 12 and 13 were obtained in good (80%) to excellent (90%) yields, independently of the chain length. Precursors 12 and 13 were then deprotected by hydrogenolysis (10% Pd/C, EtOH, 1 M HCl) to afford the target derivatives 7 and 8 in almost quantitative yield.

2. Biological Evaluation
The small library of DNJ derivatives were submitted to a panel of biological evaluations, which included inhibition of glycosidases, inhibition kinetics of β-glucosidase, as well as inhibition of B10F16 cells growth. These experiments are summarized below.

2. 1. Inhibition of Glucosidases
Glycosidase inhibitory activities of compounds 7 and 8 was evaluated against α-glucosidase (yeast), β-glucosidase (almonds), α-mannosidase (jack bean), with reference to the known standard DNJ. The results were expressed as the inhibition of glucosidase activity (IC 50 ) and are summarized in Table 1.
Compounds 7a, 7b and 7c had weak inhibitory activities against α-and β-glucosidase at 1 mM. It was, however, interesting to note that compound 7d bearing a morpholine ring was the only derivative in our library exhibiting higher and selective activity of β-glucosidase with an IC 50 of 0.052 ± 0.004 mM compared to DNJ (IC 50 = 0.65 ± 0.04 mM), while none of the other glycosidases were inhibited by this compound (Table 1). This indicated that a much more favorable interaction with the β-glucosidase active site.

2. 2. Inhibition Kinetics of β-Glucosidase
In order to explore further insight into how 7d interacted with β-glucosidase (almonds), the mode of inhibition and inhibition constant of 7d was determined by the Lineweaver-Burk plots ( Figure 3). The double reciprocal plots of 7d showed straight lines with the same v max . This indicated that 7d (K i = 7 µM) inhibited β-glucosidase in a competitive manner, a nearly 7-fold increase compared to DNJ 26 (K i = 47 µM). Hence, this competitive inhibition indicated that 7d was expected to bind to the active site of β-glucosidase and compete with their primary substrates. Moreover, a probable hydrogen bond acceptor was the carbonyl hydrogen atom of the catalytic acid. 27

Conclusion
In summary, a series of DNJ derivatives were designed and synthesized, and the structures of synthesized compounds were confirmed by 1 H NMR, 13 C NMR and HRMS. Moreover, the preliminary glucosidase inhibition and anticancer activities were evaluated in vitro. Compound 7d proved to be the most potent and selective β-glucosidase inhibitor in a competitive manner, and none of the other glycosidases were inhibited by this compound at micromolar level. Compounds 8b and 8c were moderate and selective α-glucosidase inhibitors. Nevertheless, all compounds could not inhibit the growth of B16F10 melanoma cells. The collective results indicated that a lengthening of the alkyl chain linking DNJ provide better selectivity towards α-glucosidase. The size of the hydrophobic group at the alkyl chain, especially its nature, differs greatly for the selective inhibition aganist α-and β-glucosidases. Compounds 7d, 8b and 8c would be a lead for designing novel compounds, and further derivatives would be prepared by altering these specific molecules. In addition, our results provides useful clues for the design of selective glucosidase inhibitors.

Inhibition of B16F10 Cells Growth
The inhibition of B16F10 cells growth by compounds was determined using the MTT assay and the results are summarized in Figure 4. All compounds were inactive with no significant inhibition being observed at 0.05 mM and 1 mM. This indicated that compounds by the modification of changing length of the tether, the size and nature of the terminal tertiary amine substituents had no influence on the anticancer activity.