In vivo anti-diabetic and anti-lipidemic evaluations of an excellent synthetic α-glucosidase inhibitor with dihydropyrano[3,2-c]quinoline skeleton

Abstract

Objectives

The in vivo assay is a key step in the development of a new bioactive compound as a lead drug structure. Based on importance of α-glucosidase inhibitors in the control of blood glucose level (BGL) in diabetes, in the present work, 3-amino-1-(4-chlorophenyl)-12-oxo-11,12-dihydro-1H-benzo[h]pyrano[3,2-c]quinoline-2-carbonitrile (ACODDHBPQC) that showed excellent inhibitory activity on the yeast form of α-glucosidase was selected for in vivo anti-diabetic assay.

Methods

The in vivo anti-diabetic and anti-lipidemic effects of this synthetic compound were evaluated using by a streptozotocin (STZ)-induced diabetic Wistar rat model. In silico docking study of ACODDHBPQC was performed by Atodock tools and absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties of this compound was predicted by PreADMT online software.

Results

The obtained results revealed that selected compound ACODDHBPQC showed a significant anti-diabetic effect on diabetic rats. In vivo anti-lipidemic assay also demonstrated that ACODDHBPQC had favorable effects on cholesterol and LDL levels. Furthermore, in silico studies showed that ACODDHBPQC interacted with key residues of the α-glucosidase active site and had good pharmacokinetic and toxicity properties.

Conclusion

In summary, anti-hyperglycemic effects of ACODDHBPQC was confirmed by in vivo study. However, more evaluations are needed to introduce ACODDHBPQC as a lead drug compound.

Graphical abstract

Introduction

Diabetes mellitus (DM) is the most common metabolic disorder that has two main type: type 1 (T1DM) and type 2 (T2DM) [1]. T1DM is insulin-dependent and controlled with injection of insulin, but the T2DM which is the most common type with 90% occurrence, is not insulin-dependent and controlled with oral blood glucose level (BGL) decreasing drugs [2–4]. Inhibiting the absorption of food carbohydrates is an important approach in the control of BGL in T2DM [5]. The breakdown of carbohydrates begins in saliva and reaches its peak in the small intestine [6]. The most important enzymes in the degradation of carbohydrates are α-amylase and α-glucosidase [7]. α-Amylase in saliva and pancreas and α-glucosidase in small intestine are found [8–10]. α-Amylase degraded polysaccharides to oligosaccharides and disaccharides and α-glucosidase degraded the latter carbohydrates to monosaccharaides such as glucose (Fig. 1) [11]. The result of the action of α-amylase and α-glucosidase is an increase in postprandial hyperglycemia (PPH) due to the release of glucose into the bloodstream [12]. Therefore, inhibition of these enzymes can delay the absorption of glucose into the bloodstream and lead to decrease PPH [13]. The latter procedure is an important mechanism in the treatment of T2DM. Due to the fact that inhibiting α-amylase leads to increase undigested starch secretion into the large intestine and occurrence of gastrointestinal complications, inhibition of α-glucosidase has become a more important target in the design of new anti-diabetic agents [14].

Fig. 1.

Hydrolysis of polysaccharides to simple carbohydrates by α-amylase and α-glucosidase

Considering the importance of α-glucosidase inhibitors in controlling diabetes, numerous research groups are working on the introduction of natural, semi-synthetic, and synthetic compounds as inhibitors of this enzyme [15–17]. In this regards, our research group also introduced several series of derivatives that show significant inhibition effects against α-glucosidase [18–20]. For example, as can be seen in Fig. 2, dihydropyrano[3,2-c]quinoline derivatives A demonstrated high anti-α-glucosidase activity [21]. General scaffold of compounds A, synthesized derivatives of this scaffold, range of anti-α-glucosidase activity, and the most potent compound among them are showed in Fig. 2. As seen in this figure, among the compounds A, the best derivative was 3-amino-1-(4-chlorophenyl)-12-oxo-11,12-dihydro-1H-benzo[h]pyrano[3,2-c]quinoline-2-carbonitrile (ACODDHBPQC). This compound was 72.81 fold more potent than acarbose (positive control). Furthermore, in vitro cytotoxic study of ACODDHBPQC also demonstrated that this compound was a safe compound [21]. Therefore, these in vitro results encouraged us to in vivo evaluation of ACODDHBPQC on the diabetic rats in the present work.

Fig. 2.

General structure, constructed derivatives, anti-α-glucosidase activity range, and the most potent compound among the reported derivatives A

Materials and methods

The ethics code for this work is IR.IAU.AMOL.REC.1400.002. STZ and acarbose were procured from Santa Cruz (USA) and Sigma-Aldrich CO (USA), respectively.

Chemistry

ACODDHBPQC was synthesized based on our previous work [21]. For the synthesis of this compound, a mixture of 1-naphthyl amine (2 mmol), malonic acid (4 mL), and polyphosphoric acid (1 g) was heated at 160 °C for 30 min. After that, the temperature was increased to 190 °C and the reaction continued for 2 h. The obtained mixture was cooled down to room temperature (RT). A saturated solution of sodium hydroxide was prepared and added to the mixture. This mixture was left overnight [21]. After that, a precipitate was formed that after filtration was acidified with HCl (pH = 3). The final precipitate was filtered, washed with cold water, and dried at RT to afford 4-hydroxybenzo[h]quinolin-2(1H)-one. In the next step, 4-hydroxybenzo[h]quinolin-2(1H)-one (1 mmol), 4-chlorobenzaldehyde (1.1 mmol), malononitrile (1.1 mmol), L-proline (20 mol%), and ethanol (5 mL) were poured in a sealed tube and this tube was heated at 120 °C for 3 h. The formed participate was filtered and washed with cold ethanol and water. The residue was dried at RT to afford pure product ACODDHBPQC.

In vivo biological assays

Animals

Adult male Wistar rats (n = 30; 200–250 g) were prepared from the Babol University of Medical Sciences Animal Care Center (Iran). Rats were maintained under a standard controlled conditions (temperature of 25 ºC, 12 h light–12 h dark cycle, adequate water, and food). Acarbose and ACODDHBPQC were administered orally to the target rats.

Experimental design

The wistar rats were randomly divided into five groups of 6 rats and were monitored for 30 day:

1. Group 1: Non-diabetic rats that received no treatment.

2. Group 2: Diabetic rats that received no treatment.

3. Group 3: Diabetic rats that were treated with acarbose at dose 20 mg/kg.

4. Group 4: Diabetic rats that were treated with ACODDHBPQC at dose 10 mg/kg.

5. Group 5: Diabetic rats that were treated with ACODDHBPQC at dose 20 mg/kg.

Induction of diabetes in the test animals was performed by a standard method [22]. In used protocol, STZ was injected by intracardiac route at a dosage of 45 mg/kg. After a 72 h period post-STZ injection, BGL was assessed. Wistar rats with BGL higher than 250 mg/dl were considered as diabetic rats and used for subsequent experiments. After ensuring the induction of diabetes in wistar rats, within 30 days, the acarbose and ACODDHBPQC were suspended in water and fed to the diabetic rats by gavage every day for once. At the end of the 30th day, the BGL in the studied rates after 8 h food deprivation was measured.

Biochemical assessment

BGLs were obtained by a glucometer (On Call Plus, China) and cholesterol, HDL, triglyceride, and LDL were determined by biochemical assay kits (Pars Azmoon, Iran) with a sensitivity of 0.1 mg/dl.

Statistical analysis

Statistical comparisons among the studied different groups and evaluations of the observed results were conducted using SPSS software. One-way analysis of variance (ANOVA) was employed, followed by a post-hoc Tukey HSD test, with significance set at p < 0.05.

In silico studies

Molecular modeling

Molecular modeling of ACODDHBPQC was conducted using by AutoDock Tools 1.5.6. The crystal structure of a human α-glucosidase (PDB ID: 2QMJ) as a complex bound with acarbose, was taken from RCSB PDB (http://www.rcsb.org) [23]. Subsequently, additional molecules (water molecules and original inhibitors) were removed from the enzyme structure, and the rest file was converted to a pdbqt file using by AutoDock Tools. The 3D structure of ACODDHBPQC was constructed using MarvinSketch 5.10.4, 2012, ChemAxon (http://www.chemaxon.com) and converted to a pdbqt file by AutoDock Tools. The active site coordinates of the target protein based on the position of original inhibitor acarbose in the crystal structure of α-glucosidase was identified by BIOVIA Discovery Studio 3.5, 2019 software as follows:

• Center of the grid box: x = -19.4825, y = -6.82, and z = -9.047 Å

• Dimensions of the active site box: 36 × 36 × 36 Å

The constructed pdbqt files of enzyme and ligand were used as input files for the AUTOGRID program. Flexible ligand docking was performed for ACODDHBPQC with 50 runs. The best pose of the target ligand was selected for analyzing the interactions between α-glucosidase and the latter ligand. The results were visualized using BIOVIA Discovery Studio.

Prediction of absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties

The in silico ADMET prediction for acarbose and ACODDHBPQC was conducted using the preADMET online server [24].

Results

Synthesis of ACODDHBPQC

The synthetic pathway to obtain ACODDHBPQC is outlined in Scheme 1. Initially, 4-hydroxybenzo[h]quinolin-2(1H)-one was prepared by a reaction of 1-naphthyl amine and malonic acid in the presence of polyphosphoric acid. Then, 4-hydroxybenzo[h]quinolin-2(1H)-one reacted with 4-chlorobenzaldehyde and malononitrile using L-proline in EtOH to produce ACODDHBPQC in good yields (81%) [21].

Scheme 1.

Synthesis of ACODDHBPQC. Reagents and conditions: a Polyphosphoric acid, 190 °C, 3 h; b L-proline, ethanol, 120 °C, 8 h

In vivo anti-diabetic assay

In vivo studies were performed on five groups of Wistar rats: (1) normal group (non-diabetic), (2) untreated diabetic group (diabetic), (3) diabetic group treated with acarbose at a dose of 20 mg/kg, (4) diabetic group treated with ACODDHBPQC at a dose of 10 mg/kg, and (5) diabetic group treated with ACODDHBPQC at a dose of 20 mg/kg [25]. The average BGL in the groups 2–5 were 399.33, 551.75, 395.11, and 280.32 mg/dl, respectively. Therefore, the rats were successfully diabetic (BGL > 250 mg/dL). After thirty days, the results of BGL measurement are shown in Fig. 3. The summary of the obtained results in the diabetic Wistar rats is as follows:

1. The average of BGL in group 2 increased from 399.33 mg/dl to 516 mg/dl (P = 0.006).

2. The administration of acarbose (20 mg/kg) to group 3 led to a significant decrease in the average of BGL from 551.75 mg/dl to 291.25 mg/dl (P = 0.016).

3. The administration of ACODDHBPQC (10 mg/kg) to group 4 decreased the average BGL from 395.11 mg/dl to 181.50 mg/dl (P = 0.003).

4. The administration of ACODDHBPQC (20 mg/kg) to group 5 led to in a significant decrease in the average BGL from 280.32 mg/dl to 136.67 mg/dl (P = 0.003).

Fig. 3.

Diagram of effect of acarbose and ACODDHBPQC on BGL in STZ-diabetic Wistar rats

Lipid profile evaluations

On the 30th day of treatment, the lipid profile of the studied diabetic rats was evaluated. In this evaluation, cholesterol, triglyceride, HDL, and LDL were measured in the groups 1–5 (Fig. 4).

Fig. 4.

Effect of acarbose and ACODDHBPQC on cholesterol, triglyceride, HDL, and LDL in the studied animals

The following results were obtained (Fig. 4):

1. The lowest cholesterol level was observed in group 5 (61.45 mg/dl), while the highest level was observed in group 3 (106.33 mg/dl).

2. The lowest triglyceride level was observed in group 3 (37.66 mg/dl), while the highest level was observed in group 4 (132.12 mg/dl).

3. The highest level of HDL was recorded in group 3 (67.66 mg/dl), while the lowest HDL levels were observed in group 5 (34.63 mg/dl).

4. The lowest LDL level was observed in groups 4 and 5, while the highest LDL level was recorded in group 3 (39.61 mg/dl).

Docking study

By AutoDock Tools, ACODDHBPQC was docked in the active site of the human α-glucosidase. Superimposed structure of acarbose (attached inhibitor to enzyme) and ACODDHBPQC showed in Fig. 5. Furthermore, details of interaction modes of acarbose and ACODDHBPQC showed in Fig. 6.

Fig. 5.

Acarbose (cyan) and ACODDHBPQC (pink) superimposed in the active site pocket of the human α-glucosidase

Fig. 6.

Interaction modes of acarbose (a) and ACODDHBPQC (b) in the active site of human α-glucosidase

As can be seen Fig. 6, acarbose established the fallowing interactions with the ten amino asides: hydrogen bonds with residues Asp542, Arg526, Met444, Asp327, Asp203, and His600, hydrophobic interactions with residues Phe575, Trp406, and Tyr299, and a non-classical hydrogen bond with residue Thr205.

The studied compound ACODDHBPQC established two hydrogen bonds with residues Met444 and Arg526 (Fig. 6). ACODDHBPQC also formed the fallowing π-interactions with the active site: three π-sulfur interactions with Met444, two π-anion interactions with Asp542, a π-anion interaction with Asp203, two π-π interactions with Tyr299, and two π-π interactions with Trp406. Furthermore, this compound created several hydrophobic interactions with Lys480, Ile328, and Phe450.

ADMET prediction

ADMET prediction for ACODDHBPQC and acarbose was performed using by the PreADMET online software. Results indicated that the title compounds had poor permeability to Caco-2 cells (Table 1). Moreover, both studied compounds had permeability to the blood-brain barrier (BBB) but ACODDHBPQC demonstrated high human intestinal absorption (HIA) while acarbose had not HIA. In term of toxicity: (1) both ACODDHBPQC and acarbose were mutagen, (2) ACODDHBPQC had not carcinogen on rat and mice while acarbose had a carcinogenic effect on mice, and (3) the cardiotoxicity (hERG inhibition) of ACODDHBPQC had medium risk while the cardiotoxicity of acarbose was ambiguous.

Table 1.

ADMET prediction of ACODDHBPQC and acarbose

Pharmacokinetic properties	ACODDHBPQC	Acarbose
Caco-2 cell permeability	21.1635	9.44448
HIA	96.489300	0.000000
BBB permeability	0.169689	0.0271005
Ames test	Mutagen	Mutagen
Carcino mouse	Negative	Positive
Carcino rat	Negative	Negative
hERG inhibition	Medium risk	Ambiguous

^aThe recommended ranges for Caco2: <25 poor, > 500 great, HIA: >80% is high < 25% is poor, BBB = − 3.0–1.2

Discussion

Inhibition of α-glucosidase as an important carbohydrate hydrolyzing enzyme is a main strategy for controlling of BGL in T2DM and treatment of obesity [26]. The main importance of inhibiting this enzyme is that the food carbohydrates are excreted from the digestive system without being broken down [7]. There are several α-glucosidase inhibitors in the pharmaceutical market, which are known as high consumption drugs for the control of T2DM [27]. Because the use of these drugs is associated with unfavorable gastrointestinal side effects, the design of new inhibitors for α-glucosidase is still an attractive gold for medicinal chemists [28]. At the moment, our research group is a diligent group in designing new synthetic α-glucosidase inhibitors by considering active pharmacophores [18–20]. Based on the drug development principles, according to a regular program, after in vitro evaluations, we decided to select the best α-glucosidase inhibitors of each reported series and their in vivo evaluations [29].

One of the effective pharmacophors in designing potent α-glucosidase inhibitors is pyrano[3,2-c]quinolone skeleton [30]. We synthesized derivatives of this skeleton and after in vitro evaluations, we arrive at a derivative with abbreviated name of ACODDHBPQC) that can inhibit α-glucosidase 72.81 times better than acarbose [21]. Therefore, in vivo evaluation of this derivative could be very valuable in designing a lead compound to reach a new drug.

In vivo anti-diabetic evaluations in the present work revealed that ACODDHBPQC in two doses 10 and 20 mg/kg clearly reduced BGL. Furthermore, in silico docking study on ACODDHBPQC demonstrated that this compound formed eight important interactions with the α-glucosidase active site: three π-sulfur and one hydrogen bond with Met444, one hydrogen bond with Arg526, two π-anion interactions with Asp542, and one π-anion interaction with Asp203. These four latter amino acids (Met444, Arg526, Asp542, and Asp203) also observed in the binding mode of the standard inhibitor acarbose. On the other hand, interactions with the Asp203 and Asp542 have been observed in the synthetic compounds with significant in vivo anti-diabetic effect [31]. Therefore, it seems that ACODDHBPQC by establishing interactions with important amino acids of the α-glucosidase active site created anti-diabetic effect.

Based on the importance of the lipid profile in patients with T2DM, we evaluated the effects of ACODDHBPQC on this profile in diabetic rats and obtained results compared with acarbose [32]. The obtained results demonstrated that ACODDHBPQC reduced cholesterol and LDL levels while acarbose significantly increased HDL level and decreased triglyceride level. Therefore, effect of ACODDHBPQC on the lipid profile is appropriate. Furthermore, in silico ADMET study demonstrated that ACODDHBPQC had pharmacokinetic properties comparable with acarbose and toxicity profile better than acarbose. However, it should be noted that HIA of ACODDHBPQC is predicted to be high, but acarbose has not HIA. Considering that acarbose does not require intestinal absorption to produce its effect, this case does not give a special advantage to our synthetic compound.

Conclusion

In the present study, ACODDHBPQC as a dihydropyrano[3,2-c]quinolone derivative that showed high α-glucosidase inhibition in in vitro study was selected for in vivo anti-diabetic and anti-lipidemic evaluations. Anti-diabetic study on STZ-induced diabetic Wistar rats demonstrated that ACODDHBPQC significantly reduced BGL. Moreover, in vivo anti-lipidemic evaluations revealed that ACODDHBPQC reduced cholesterol and LDL levels and had not significant effects on triglyceride and HDL levels in diabetic Wistar rats. Furthermore, the docking study of ACODDHBPQC on the active site of human α-glucosidase demonstrated that this compound interacted with the key residues of active site. Moreover, ADMET study predicted that ACODDHBPQC had appropriate toxicity profile and could be considered as a potential oral drug candidate. In conclusion, this work provides valuable insights related to the anti-diabetic and anti-lipidemic effects of ACODDHBPQC as a promising candidate for further drug development.

Abbreviations

ACODDHBPQC

3-Amino-1-(4-chlorophenyl)-12-oxo-11,12-dihydro-1H-benzo[h]pyrano[3,2-c]quinoline-2-carbonitrile

ADMET

Absorption, Distribution, Metabolism, Excretion, and toxicity

BBB

Blood brain barrier

BGL

Blood glucose level

DM

Diabetes mellitus

HIA

Human intestinal absorption

PPH

Postprandial hyperglycemia

RT

Room temperature

STZ

Streptozotocin

T1DM

Type 1 diabetes mellitus

T2DM

Type 2 diabetes mellitus

Funding

None.

Data availability

The data that support the findings of this study are available on request from the corresponding author.

Declarations

Ethical approval

The ethics code for this work is IR.IAU.AMOL.REC.1400.002.

Conflict of interest

The authors declared no conflict of interest in this study.