An efficient and targeted synthetic approach towards new highly substituted 6-amino-pyrazolo[1,5-a]pyrimidines with α-glucosidase inhibitory activity

In an attempt to find novel α-glucosidase inhibitors, an efficient, straightforward reaction to synthesize a library of fully substituted 6-amino-pyrazolo[1,5-a]pyrimidines 3 has been investigated. Heating a mixture of α-azidochalcones 1 and 3-aminopyrazoles 2 under the mild condition afforded desired compounds with a large substrate scope in good to excellent yields. All obtained products were evaluated as α-glucosidase inhibitors and exhibited excellent potency with IC50 values ranging from 15.2 ± 0.4 µM to 201.3 ± 4.2 µM. Among them, compound 3d was around 50-fold more potent than acarbose (IC50 = 750.0 ± 1.5 µM) as standard inhibitor. Regarding product structures, kinetic study and molecular docking were carried out for two of the most potent ones.

In vitro α-glucosidase inhibitory activity. The obtained highly substituted 6-amino-pyrazolo [1,5-a] pyrimidines 3 were evaluated for their in vitro inhibitory activities against α-glucosidase (Saccharomyces cerevisiae, EC.3.2.1.20) and the results were compared with acarbose as the reference drug (Tables 1 and 2). As it can be seen, all the synthesized compounds showed good to excellent inhibitory activities with IC 50 values of 15.2 ± 0.4−201.3 ± 4.2 µM in comparison to the standard drug IC 50 = 750.0 ± 1.5 µM. To explain the structure and observed activity correlations, the 6-amino-pyrazolo[1,5-a]pyrimidines 3 were divided into two categories based on the substituents on the pyrazole moiety: the presence of amide functional group at C3-position 3a-z (summarized in Table 1) along with ester functional group at C3-position 3aa-ai (summarized in Table 2). Additionally, the substituents on the 5-phenyl and 7-phenyl rings of pyrimidine ring were changed in each series to optimize the α-glucosidase inhibition.
Among the 6-amino-pyrazolo [1,5-a]pyrimidines 3a−i, compound 3d with 4-CH 3 substitutent on the 5-aryl and 4-Br substituent on the 7-aryl ring showed the most inhibitory activity in this series (IC 50 = 15.2 ± 0.4 µM). It is worth mentioning that this derivative showed the highest anti-α-glucosidase potency among all the synthesized compounds. Removal of the methyl group from 5-phenyl ring (compounds 3a and 3b) and also replacement of bromine with chlorine atom (compound 3c) led to the significant decrease in inhibitory activity. 4-OCH 3 substituent on the 5-phenyl ring resulted into a considerable deterioration in activity (compounds 3e and 3f). Compound 3g with 4-Cl substituent on the 7-phenyl ring showed low activity (IC 50 = 75.6 ± 5.0). Adding a chlorine atom to 5-position of phenyl ring, or replacing it with a heterocycle caused very good effect on the observed activities (compounds 3h and 3i).
In the second series, the 4-OCH 3 substituted N-phenyl-pyrazolo-5-carboxamides 3j-r, compound 3m, which is the analog of compound 3d, showed the best activity against α-glucosidase. There was the same trend for the activities of compounds 3j-l with their analog in the first series. Compounds 3n and 3o with 4-OCH 3 substituted 5-phenyl ring showed a moderate activity. The replacement of methoxy group with chlorine atom at 4-position of 5-phenyl ring led to the better performance (compounds 3p and 3q). Introducing a heterocycle on the 7-phenyl ring has a good effect to increase activity (compound 3r).
Among the synthesized derivatives in third series, compound 3v was found to be the most potent compound. Same as two previous series, removal of methyl group or addition of chlorine atom (compounds 3s-v) had a destructive effect on the observed inhibitory activities. It was found that introduction an electron-donating group (OCH 3 ) on the 4-postion of 5-phenyl ring (compound 3w) causes a decrease in activity against α-glucosidase (IC 50 = 53.7 ± 3.4). Finally, 4-Cl substituted 5-phenyl ring derivatives 3x-z were investigated. Compound 3x with unsubstituted 7-phenyl ring showed a weak inhibitory activity (IC 50 = 80.3 ± 5.2). Introducing another chlorine atom to 4-position of this ring (compound 3y), or thiophene (compound 3z) led to a significant increase in inhibitory activity.
In the second category, the compound 3af with 4-OCH 3 and 4-Cl substituents on the respectively 5 and 7-phenyl rings showed the highest potency against the α-glucosidase. Further changes on this compound like removing and replacing 4-OCH 3 with 4-CH 3 and 4-Cl on the 5-phenyl ring (compounds 3aa, 3ac and 3ah) as well as removing chlorine from 7-phenyl rings (compound 3ae) made notable increase in IC 50 value. Compound 3ag with 4-Cl on the 5-phenyl ring was the weakest compound in this series. Addition of another chlorine atom to 4-position of this ring (compound 3ah), or thiophene (compound 3ai) improved the inhibition activities.
Thorough the comparison of IC 50 values of synthesized3a-z with their analog 3aa-ai, it can be found that substituents on the pyrazole moiety played a substantial role on the observed α-glucosidase inhibitory activities. Although the presence of 4-OCH 3 on the 5-phenyl ring had destructive effect in the first category, the compounds containing this group showed the highest activities in the second category.
Enzyme kinetic studies. The inhibition mode of the synthesized compounds 3 against α-glucosidase was investigated. For this purpose, kinetics analysis was carried out with reference drug, acarbose, and the most potent derivative in each category (3d and 3af). The inhibition type was indicated on the basis of Michaelis-Menten and Lineweaver-Burk plots. As it can be seen in the Lineweaver-Burk plot of selected compounds (Fig. 2), with increasing inhibitor concentrations, the K m value gradually increased while V max value remained unchanged which indicated competitive inhibition. Accordingly, this study revealed both 3d and 3af compete with acarbose for binding to the enzyme active site. Furthermore, plot of the K m versus different concentration of inhibitor gave an estimate of the inhibition constant, K i of 12 µM and 65 µM for compounds 3d and 3af, respectively.
In-silico ADME evaluation. The ADME properties for some of the synthesized highly substituted 6-amino-pyrazolo[1,5-a]pyrimidines 3 were computed using Swiss ADME online (http://www.swissadme.ch/ index.php) toolkit 76 . Through this in-Silico study under the Lipniski's rule, five determined drug-likeness parameters were compared with the known drugs 77 . These evaluated parameters are summarized in Table 3. On the basis of MW (<500), HBA (≤10), HBD (<5), and log P (<5) values, the good oral bioavailability of the selected compounds can be estimated. Lipophilicity is determined by Log P in which P is the octanol-water partition coefficient. As it can be seen in Table 3, all the studied compounds have the Log P values in the desirable range. The molecular flexibility can be proved regarding the number of rotatable bonds which should be less than 10 (nROTB <10) and regarding to Table 3, all the obtained numbers are 6 and 7. Topological polar surface area (TPSA) can reveal the surface contribution of polar fragments. The high value of TPSA (>140 Å 2 ) may show low blood-brain barrier (BBB) penetration, and therefore, poor membrane permeability 78 . As it can be seen in Table 3 Acarbose ---750.0 ± 1.5 Table 1. Substrate scope and in vitro α-glucosidase inhibitory activity of compounds3a-z. a Values are the mean ± SD. All experiments were performed at least three times.
Dock Tools (version 1.5.6) to study ligand-enzyme interactions 21,80 . Briefly, crystal structures of isomaltase from Saccharomyces cerevisiae (PDB code 3A4A), with 72% identical and shares 85% similarity with the Saccharomyces cerevisiae α-glucosidase, was designated for building modeled α-glucosidase. Afterward, the interaction modes of acarbose as standard inhibitor and the most potent compound in each category 3d and 3af in the active site of α-glucosidase were studied. As shown in Fig. 3, acarbose formed interactions with Asn241, His279, Glu304, Arg312, Thr302, Thr307, Ser308, and Gln322 residues in the enzyme active site. For the most active compound 3d, amino group established hydrogen bonds with active site residues Thr307 and Glu304. Furthermore, 4-bromo phenyl moiety formed a π-anion interaction with Glu304. Several hydrophobic interactions were also observed with the active site residues His239, Pro309, Arg312 and Ala326. The interactions of 3d are shown in Fig. 4a. In the case of 3af (Fig. 4b), hydrogen bonds between amino group and the active site residues Thr307 and Glu304 were formed. Glu304 interacted with 4-choloro phenyl moiety to form π-anion interaction. In addition, several hydrophobic interactions were observed between His239, Val305, Pro309, and Arg312 and 4-choloro phenyl moiety. Further studies on binding energies of compounds 3d, 3af and acarbose revealed that they have lower free binding energy (3d: −10.0 kcal/mol and 3af: −9.57 kcal/mol) than acarbose (−4.04 kcal/mol). This means they can bond easier to the target enzyme in comparison to acarbose.

Conclusion.
In conclusion, we have represented a Michael-addition-cyclocondensation reaction between α-azidochalcones and 3-aminopyrazoles to prepare a novel library of fully substituted pyrazolo[1,5-a]pyrimidines and evaluated their α-glucosidase activities. Providing an efficient, simple protocol from readily available starting materials, this method led to new 6-amino-pyrazolo[1,5-a]pyrimidines in short time and under the mild conditions. Additionally, easy work-up without any need for chromatography purification processes and really good product yields are the significant features of this proposed reaction. The synthesized compounds were investigated by α-glucosidase inhibitory activity assay. All of them showed very good to excellent activities in comparison to the standard drug. Among these derivatives, 3d was the most potent one with IC 50 value of 15.2±0.4 µM. The kinetic analysis for the most active compound from each category (3d and 3af) compound  Table 2. Substrate scope and in vitro α-glucosidase inhibitory activity of compounds 3aa-ai. a Values are the mean ± SD. All experiments were performed at least three times.
www.nature.com/scientificreports www.nature.com/scientificreports/ showed that there was a competitive mechanism to inhibit α-glucosidase. Furthermore, docking studies for these products revealed there were several interactions between desired compounds and important amino acids in the active site of the enzyme.

Experimental
Methods. All chemicals were purchased from Merck (Germany) and were used without further purification.
Molecular docking studies. Autodock 4.2.6 program was used to determine the probable binding conformations of the compounds 3d and 3af over the α-glucosidase active site. AutoDockTools 1.5.2 (ADT) was utilized to prepare the input files. The 3D structure of the most active compounds were drawn 3d and 3af using MarvineSketch 5.8.3, 2012, ChemAxon (http://www.chemaxon.com) and converted to pdbqt coordinate using Auto dockTools 85 . In AUTOGRID for each atom type in the ligand, maps were calculated with 0.375 Å spacing between grid points, and a grid box of 40 × 40 × 40 Å was created at the center of 12.5825, −7.8955, 12.519 in each dimension to determine the ligand-enzyme interactions. Rigid ligand dockings were accomplished for the selected compounds. Of the three different search algorithms suggested by AutoDock 4.2.6, the Lamarckian genetic algorithm (LGA) consisting of 50 runs, 25 × 10 6 energy evaluations, and 27,000 generations was carried out 86 . Other docking parameters were set to default. The best interaction of the selected compounds were considered for analyzing and the results were illustrated using Discovery Studio 4.5 Client.