α-Glucosidase and α-Amylase Inhibitory Potentials of Quinoline–1,3,4-oxadiazole Conjugates Bearing 1,2,3-Triazole with Antioxidant Activity, Kinetic Studies, and Computational Validation

Diabetes mellitus (DM) is a multifaceted metabolic disorder that remains a major threat to global health security. Sadly, the clinical relevance of available drugs is burdened with an upsurge in adverse effects; hence, inhibiting the carbohydrate-hydrolyzing enzymes α-glucosidase and α-amylase while preventing oxidative stress is deemed a practicable strategy for regulating postprandial glucose levels in DM patients. We report herein the α-glucosidase and α-amylase inhibition and antioxidant profile of quinoline hybrids 4a–t and 12a–t bearing 1,3,4-oxadiazole and 1,2,3-triazole cores, respectively. Overall, compound 4i with a bromopentyl sidechain exhibited the strongest α-glucosidase inhibition (IC50 = 15.85 µM) relative to reference drug acarbose (IC50 = 17.85 µM) and the best antioxidant profile in FRAP, DPPH, and NO scavenging assays. Compounds 4a and 12g also emerged as the most potent NO scavengers (IC50 = 2.67 and 3.01 µM, respectively) compared to gallic acid (IC50 = 728.68 µM), while notable α-glucosidase inhibition was observed for p-fluorobenzyl compound 4k (IC50 = 23.69 µM) and phenyl-1,2,3-triazolyl compound 12k (IC50 = 22.47 µM). Moreover, kinetic studies established the mode of α-glucosidase inhibition as non-competitive, thus classifying the quinoline hybrids as allosteric inhibitors. Molecular docking and molecular dynamics simulations then provided insights into the protein–ligand interaction profile and the stable complexation of promising hybrids at the allosteric site of α-glucosidase. These results showcase these compounds as worthy scaffolds for developing more potent α-glucosidase inhibitors with antioxidant activity for effective DM management.


Introduction
Diabetes Mellitus (DM) is a chronic systemic disorder characterized by dysfunctional glucose regulation. Typically, patients secrete insufficient insulin (i.e., type 1 diabetes mellitus-T1DM) or develop insulin resistance (i.e., type 2 diabetes mellitus-T2DM), which leads to elevated blood glucose levels (hyperglycemia) [1]. Complications associated with DM include nervous system damage, eye damage, cardiovascular disease, stroke, kidney disease, and gangrene [2]. The disease persists as a heavy burden on public health, having become the ninth leading cause of death with approximately 451 million incidences globally [3]. Recent estimations have shown increased incidences in Africa, with about 15.5 million cases in 2017, and that 69.2% of the adult population are ignorant of their diabetic status [4]. T2DM has a higher incidence rate globally, with about 90% or more of the cases relative to T1DM. T2DM can be managed by regulating the key molecular targets involved in carbohydrate digestion i.e., α-glucosidase and pancreatic α-amylase [5]. These enzymes break down dietary carbohydrates into simple sugars which are absorbed from the small intestine into the bloodstream. Elevated levels of α-glucosidase and α-amylase are therefore associated with increased postprandial glucose levels seen in T2DM patients, making their inhibition a successful strategy for T2DM management [6]. Nevertheless, the availability of α-glucosidase inhibitors is currently restricted to N-heterocyclic carbasugars such as acarbose, voglibose, and miglitol [7]. These molecules present sugar-like moieties that compete with the enzyme's natural substrates for binding at the active site to inhibit sugar hydrolysis, and consequently decrease postprandial hyperglycemia [8]. Although these drugs have rapid action, they are marred by efficacy problems and adverse side effects such as diarrhea, flatulence, and abdominal discomfort [9], which precipitates the need for new and safer α-glucosidase inhibitors.
Reactive oxygen species (ROS), including free radicals of superoxide anion (O 2 − ), hydroxyl (OH − ) radicals, and hydrogen peroxide (H 2 O 2 ), play a prominent role in regulating cellular functions as their intermediates control various enzymatic reactions involved in signal transductions [10]. However, their excessive accumulation contributes to toxic cellular oxidative stress which induces the pathogenesis of diseases such as T2DM [11]. Therefore, the development of small molecules with antidiabetic and free radical scavenging potentials is considered an attractive option for T2DM management.
Although quinoline, 1,3,4-oxadiazole, and 1,2,3-triazole moieties have been reported as α-glucosidase inhibitors in many molecular hybrid designs [29], these findings require further investigation to achieve optimal results. Consequently, in continuation of our exploits within the MH approach towards quinoline, 1,3,4-oxadiazole, and 1,2,3-triazole hybrids with therapeutic potentials [30,31], we herein incorporated these pharmacophores in a single molecular hybrid and evaluated the effect on the in vitro α-glucosidase and α-amylase inhibitory potencies and antioxidant activities. We also examined the mode of α-glucosidase inhibition of promising compounds via kinetic studies, then rationalized the results with molecular docking and molecular dynamics simulations. Although quinoline, 1,3,4-oxadiazole, and 1,2,3-triazole moieties have been reported as α-glucosidase inhibitors in many molecular hybrid designs [29], these findings require further investigation to achieve optimal results. Consequently, in continuation of our exploits within the MH approach towards quinoline, 1,3,4-oxadiazole, and 1,2,3-triazole hybrids with therapeutic potentials [30,31], we herein incorporated these pharmacophores in a single molecular hybrid and evaluated the effect on the in vitro α-glucosidase and αamylase inhibitory potencies and antioxidant activities. We also examined the mode of αglucosidase inhibition of promising compounds via kinetic studies, then rationalized the results with molecular docking and molecular dynamics simulations.

Antioxidant Activity Profiling
The DPPH assay is used conventionally to quantify a compound's ability to trap free radicals. The relative efficacy of compounds 4a-t and 12a-t was evaluated in this regard; the results are shown in Table 1. Except for the hybrids' precursor, compound 3 (IC50 of 25.15 µM), the compounds in both series exhibited moderate free radical scavenging activity relative to the reference scavenger, gallic acid (IC50 = 26.63 µM). The 1,3,4-oxadiazole series 4a-t were again more potent than the 1,2,3-triazoles 12a-t. SAR analysis showed that 2-mercapto-1,3,4-oxadiazole core was beneficial to DPPH scavenging activity, as seen in the two-fold improved potency of compound 3 as compared to 8-HQ (IC50 = 51.81 µM); however, S-alkylation was found generally detrimental to potency. "-" = greater than 500 µM.

Ferric Reducing Antioxidant Power (FRAP)
FRAP is a ferric (Fe 3+ ) to ferrous (Fe 2+ ) ions reduction assay that measures a compound's electron-donating capacity, i.e., the antioxidant's reducing power [32]. The results in Table 2 reveal that compound 4i was the most promising hybrid in this regard;

Mode of α-Glucosidase Inhibition: Enzyme Kinetic Studies
Understanding a drug's mode of inhibiting its target is crucial to improve the drug's activity profile and target specificity. Although orthosteric inhibition, in which the ligand binds at the target's active site, dominates among currently available drugs, allosteric inhibitors (i.e., drugs binding at regions close to the active site) are desirable due to their reduced vulnerability to active site mutation or to displacement through native substrate overload at the active site [34]. Accordingly, the mode of α-glucosidase inhibition was investigated for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, using a time-dependent para-nitrophenyl-β-D-glucopyranoside (pNPG) assay; the results are presented as Lineweaver-Burk plots (Figure 4). Clearly, all the compounds exhibited a non-competitive type of enzyme inhibition as the maximum reaction velocity (Vmax) changed for each inhibitor concentration tested, but the Michaelis-Menten constant (Km) remained unchanged. This indicates that the present compounds bind at the allosteric sites of free α-glucosidase, to inhibit its hydrolytic activity, and of the enzyme-substrate complex, to induce a slower release of the hydrolysis product glucose. Conceivably, this can result in decreased systemic glucose concentration thus alleviating postprandial hyperglycemia and its associated complications.

Mode of α-Glucosidase Inhibition: Enzyme Kinetic Studies
Understanding a drug's mode of inhibiting its target is crucial to improve the drug's activity profile and target specificity. Although orthosteric inhibition, in which the ligand binds at the target's active site, dominates among currently available drugs, allosteric inhibitors (i.e., drugs binding at regions close to the active site) are desirable due to their reduced vulnerability to active site mutation or to displacement through native substrate overload at the active site [34]. Accordingly, the mode of α-glucosidase inhibition was investigated for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, using a time-dependent para-nitrophenyl-β-D-glucopyranoside (pNPG) assay; the results are presented as Lineweaver-Burk plots (Figure 4). Clearly, all the compounds exhibited a non-competitive type of enzyme inhibition as the maximum reaction velocity (Vmax) changed for each inhibitor concentration tested, but the Michaelis-Menten constant (Km) remained unchanged. This indicates that the present compounds bind at the allosteric sites of free α-glucosidase, to inhibit its hydrolytic activity, and of the enzyme-substrate complex, to induce a slower release of the hydrolysis product glucose. Conceivably, this can result in decreased systemic glucose concentration thus alleviating postprandial hyperglycemia and its associated complications.

Mode of α-Glucosidase Inhibition: Enzyme Kinetic Studies
Understanding a drug's mode of inhibiting its target is crucial to improve the drug's activity profile and target specificity. Although orthosteric inhibition, in which the ligand binds at the target's active site, dominates among currently available drugs, allosteric inhibitors (i.e., drugs binding at regions close to the active site) are desirable due to their reduced vulnerability to active site mutation or to displacement through native substrate overload at the active site [34]. Accordingly, the mode of α-glucosidase inhibition was investigated for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, using a time-dependent para-nitrophenyl-β-D-glucopyranoside (pNPG) assay; the results are presented as Lineweaver-Burk plots (Figure 4). Clearly, all the compounds exhibited a non-competitive type of enzyme inhibition as the maximum reaction velocity (Vmax) changed for each inhibitor concentration tested, but the Michaelis-Menten constant (Km) remained unchanged. This indicates that the present compounds bind at the allosteric sites of free α-glucosidase, to inhibit its hydrolytic activity, and of the enzyme-substrate complex, to induce a slower release of the hydrolysis product glucose. Conceivably, this can result in decreased systemic glucose concentration thus alleviating postprandial hyperglycemia and its associated complications.

Mode of α-Glucosidase Inhibition: Enzyme Kinetic Studies
Understanding a drug's mode of inhibiting its target is crucial to improve the drug's activity profile and target specificity. Although orthosteric inhibition, in which the ligand binds at the target's active site, dominates among currently available drugs, allosteric inhibitors (i.e., drugs binding at regions close to the active site) are desirable due to their reduced vulnerability to active site mutation or to displacement through native substrate overload at the active site [34]. Accordingly, the mode of α-glucosidase inhibition was investigated for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, using a time-dependent para-nitrophenyl-β-D-glucopyranoside (pNPG) assay; the results are presented as Lineweaver-Burk plots (Figure 4). Clearly, all the compounds exhibited a non-competitive type of enzyme inhibition as the maximum reaction velocity (Vmax) changed for each inhibitor concentration tested, but the Michaelis-Menten constant (Km) remained unchanged. This indicates that the present compounds bind at the allosteric sites of free α-glucosidase, to inhibit its hydrolytic activity, and of the enzyme-substrate complex, to induce a slower release of the hydrolysis product glucose. Conceivably, this can result in decreased systemic glucose concentration thus alleviating postprandial hyperglycemia and its associated complications.

Mode of α-Glucosidase Inhibition: Enzyme Kinetic Studies
Understanding a drug's mode of inhibiting its target is crucial to improve the drug's activity profile and target specificity. Although orthosteric inhibition, in which the ligand binds at the target's active site, dominates among currently available drugs, allosteric inhibitors (i.e., drugs binding at regions close to the active site) are desirable due to their reduced vulnerability to active site mutation or to displacement through native substrate overload at the active site [34]. Accordingly, the mode of α-glucosidase inhibition was investigated for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, using a time-dependent para-nitrophenyl-β-D-glucopyranoside (pNPG) assay; the results are presented as Lineweaver-Burk plots (Figure 4). Clearly, all the compounds exhibited a non-competitive type of enzyme inhibition as the maximum reaction velocity (Vmax) changed for each inhibitor concentration tested, but the Michaelis-Menten constant (Km) remained unchanged. This indicates that the present compounds bind at the allosteric sites of free α-glucosidase, to inhibit its hydrolytic activity, and of the enzyme-substrate complex, to induce a slower release of the hydrolysis product glucose. Conceivably, this can result in decreased systemic glucose concentration thus alleviating postprandial hyperglycemia and its associated complications.

Mode of α-Glucosidase Inhibition: Enzyme Kinetic Studies
Understanding a drug's mode of inhibiting its target is crucial to improve the drug's activity profile and target specificity. Although orthosteric inhibition, in which the ligand binds at the target's active site, dominates among currently available drugs, allosteric inhibitors (i.e., drugs binding at regions close to the active site) are desirable due to their reduced vulnerability to active site mutation or to displacement through native substrate overload at the active site [34]. Accordingly, the mode of α-glucosidase inhibition was investigated for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, using a time-dependent para-nitrophenyl-β-D-glucopyranoside (pNPG) assay; the results are presented as Lineweaver-Burk plots (Figure 4). Clearly, all the compounds exhibited a non-competitive type of enzyme inhibition as the maximum reaction velocity (Vmax) changed for each inhibitor concentration tested, but the Michaelis-Menten constant (Km) remained unchanged. This indicates that the present compounds bind at the allosteric sites of free α-glucosidase, to inhibit its hydrolytic activity, and of the enzyme-substrate complex, to induce a slower release of the hydrolysis product glucose. Conceivably, this can result in decreased systemic glucose concentration thus alleviating postprandial hyperglycemia and its associated complications.

Enzyme Inhibition and Structure-Activity Relationship (SAR) Analysis
Abounding empirical evidence highlights the significance in the pathogenesis of T2DM of carbohydrate-digesting enzymes α-glucosidase and α-amylase; hence, inhibiting these enzymes is crucial to T2DM management [33]. The inhibitory potencies of quinoline hybrids 4a-t and 12a-t against these enzymes are presented in Table 2. The compounds show low micromolar α-glucosidase inhibition with IC 50 values ranging between 15.85 to 63.59 µM, in contrast to their moderate α-amylase inhibition. The quinoline-1,3,4-oxadiazole series 4a-t are generally stronger α-glucosidase inhibitors compared with their 1,2,3-triazole congeners 12a-t. This SAR is conceivably due to the increased lipophilicity and polar surface area of the 1,2,3-triazole core. Interestingly, the less lipophilic phenyl compounds 12k-t were also better α-glucosidase inhibitors compared with the more lipophilic benzyl compounds 12a-j.
In the 1,3,4-oxadiazole series 4a-t, propyl was identified as the optimal alkane chain, as seen in the α-glucosidase inhibitory potencies of 4a-f. However, replacing these alkyl chains with a bromopentyl unit led to the most potent α-glucosidase inhibitor, compound 4i (IC 50 = 15.85 µM) with a superior potency compared to standard drug acarbose The trend was seen in the chloro (4l, 4p) and bromo (4m, 4q) substituted compounds, respectively. This shows that altering the substituent's position from para to meta is detrimental to α-glucosidase inhibition, and the order of potency in the para-substituted derivatives is F > Cl > Br > NO 2 . It is also inferable that inhibitory potency is dependent on the halogen's electronegativity. The non-tolerance for meta or ortho positions was again highlighted in disubstituted compounds 4q-4t which had inferior potencies relative to their monosubstituted congeners. Interestingly, those SARs were repeated in the 12a-t series. Compound 12d (IC 50 = 38.11 µM) was more potent than 12h (IC 50 = 63.59 µM) and 12i (IC 50 = 42.79 µM). Conversely, both series exhibited poor αamylase inhibition. A schematic SAR summary of the antioxidant profiling and α-glucosidase inhibition for the quinoline hybrids is presented in Figure 3.
compounds 12k-t were also better α-glucosidase inhibitors compared with the more lipophilic benzyl compounds 12a-j.
In the 1,3,4-oxadiazole series 4a-t, propyl was identified as the optimal alkane chain, as seen in the α-glucosidase inhibitory potencies of 4a-f. However, replacing these alkyl chains with a bromopentyl unit led to the most potent α-glucosidase inhibitor, compound 4i (IC50 = 15.85 µM) with a superior potency compared to standard drug acarbose (IC50 = 17.85 µM). Among the benzyl analogues, para-fluoro substituted compound 4k (IC50 = 23.69 µM) was twice as potent as the meta-fluoro analogue 4o (IC50 = 48.84 µM). The trend was seen in the chloro (4l, 4p) and bromo (4m, 4q) substituted compounds, respectively. This shows that altering the substituent's position from para to meta is detrimental to αglucosidase inhibition, and the order of potency in the para-substituted derivatives is F > Cl > Br > NO2. It is also inferable that inhibitory potency is dependent on the halogen's electronegativity. The non-tolerance for meta or ortho positions was again highlighted in disubstituted compounds 4q-4t which had inferior potencies relative to their monosubstituted congeners. Interestingly, those SARs were repeated in the 12a-t series. Compound 12d (IC50 = 38.11 µM) was more potent than 12h (IC50 = 63.59 µM) and 12i (IC50 = 42.79 µM). Conversely, both series exhibited poor α-amylase inhibition. A schematic SAR summary of the antioxidant profiling and α-glucosidase inhibition for the quinoline hybrids is presented in Figure 3.

Mode of α-Glucosidase Inhibition: Enzyme Kinetic Studies
Understanding a drug's mode of inhibiting its target is crucial to improve the drug's activity profile and target specificity. Although orthosteric inhibition, in which the ligand binds at the target's active site, dominates among currently available drugs, allosteric inhibitors (i.e., drugs binding at regions close to the active site) are desirable due to their reduced vulnerability to active site mutation or to displacement through native substrate overload at the active site [34]. Accordingly, the mode of α-glucosidase inhibition was investigated for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, using a time-dependent para-nitrophenyl-β-D-glucopyranoside (pNPG) assay; the results are presented as Lineweaver-Burk plots ( Figure 4). Clearly, all the compounds exhibited a noncompetitive type of enzyme inhibition as the maximum reaction velocity (Vmax) changed for each inhibitor concentration tested, but the Michaelis-Menten constant (Km) remained unchanged. This indicates that the present compounds bind at the allosteric sites of free α-glucosidase, to inhibit its hydrolytic activity, and of the enzyme-substrate complex, to induce a slower release of the hydrolysis product glucose. Conceivably, this can result in decreased systemic glucose concentration thus alleviating postprandial hyperglycemia and its associated complications.
itive type of enzyme inhibition as the maximum reaction velocity (Vmax) changed for each inhibitor concentration tested, but the Michaelis-Menten constant (Km) remained unchanged. This indicates that the present compounds bind at the allosteric sites of free α-glucosidase, to inhibit its hydrolytic activity, and of the enzyme-substrate complex, to induce a slower release of the hydrolysis product glucose. Conceivably, this can result in decreased systemic glucose concentration thus alleviating postprandial hyperglycemia and its associated complications.

Homology Modeling and Molecular Docking
Having established from enzyme kinetic studies the allosteric mode of α-glucosidase inhibition for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, we sought to gain an in silico insight into the compounds' binding profiles. However, the 3D-crystal structure of α-glucosidase remained unsolved; thus, we constructed a 3D model of α-glucosidase using the homology modeling module in the Schrödinger molecular modelling suite [35] and the FASTA sequence of Saccharomyces cerevisiae yeast α-glucosidase (UniProt entry: P38138) [36]. A BLAST search retuned the oligo-1,6-glucosidase isomaltase structure (PDB: 3A4A) which was selected as a template for model building. Notably, the yeast isomaltase shared a high sequence identity and similarity, 72% and 86%, respectively, with α-glucosidase; hence their 3D structures were similar ( Figure 5).
Due to the non-competitive character (i.e., allosteric inhibition) shown by the selected compounds, SiteMap calculations [37] were performed to identify other potential binding sites on α-glucosidase. The calculations produced five binding sites (Table S1) which were

Homology Modeling and Molecular Docking
Having established from enzyme kinetic studies the allosteric mode of α-glucosidase inhibition for promising compounds 4b, 4i, 4j, 4k, 4l, 12k, and 12m, we sought to gain an in silico insight into the compounds' binding profiles. However, the 3D-crystal structure of α-glucosidase remained unsolved; thus, we constructed a 3D model of α-glucosidase using the homology modeling module in the Schrödinger molecular modelling suite [35] and the FASTA sequence of Saccharomyces cerevisiae yeast α-glucosidase (UniProt entry: P38138) [36]. A BLAST search retuned the oligo-1,6-glucosidase isomaltase structure (PDB: 3A4A) which was selected as a template for model building. Notably, the yeast isomaltase shared a high sequence identity and similarity, 72% and 86%, respectively, with α-glucosidase; hence their 3D structures were similar ( Figure 5). Analysis of the α-glucosidase complexes of examined compounds ( Figure 6) revealed crucial protein-ligand interactions which substantiated the enzyme inhibition data. Foremost, the most potent compound 4i found an excellent fit at the allosteric site, with the highest docking score of −6.938 in the series reflecting its potency (Table S2). The α-glucosidase complex was characterized by hydrogen bond interactions of the quinoline nitrogen and oxygens atoms with His245, as well as a similar interaction of the oxadiazole ring's nitrogen with the amino side chain of Asn241. A face-to-face π-π stacking was present between the quinoline ring and Phe300. The ligand's stability at the allosteric site was fostered by hydrophobic interactions with Lys155, Ser156, Phe157, Phe177, His239, Pro240, His245, Ala278, and Phe311 residues. More importantly, these interactions were anchored by a halogen bond interaction of the bromopentyl unit with Agr312, Tyr313, and Asn412, an interaction which conceivably accounts for the compound's superior α-glucosidase inhibition. Due to the non-competitive character (i.e., allosteric inhibition) shown by the selected compounds, SiteMap calculations [37] were performed to identify other potential binding sites on α-glucosidase. The calculations produced five binding sites (Table S1) which were evaluated in terms of site score, drugability score (>1), and site volume (>225). Site two was chosen as the most probable allosteric site using these scoring functions (Figure 4). Subsequently, molecular docking calculations were performed at this site using the induced fit docking protocol [38], with Prime and Glide algorithms to account for the protein's conformational changes due to ligand binding.
Analysis of the α-glucosidase complexes of examined compounds ( Figure 6) revealed crucial protein-ligand interactions which substantiated the enzyme inhibition data. Foremost, the most potent compound 4i found an excellent fit at the allosteric site, with the highest docking score of −6.938 in the series reflecting its potency (Table S2). The αglucosidase complex was characterized by hydrogen bond interactions of the quinoline nitrogen and oxygens atoms with His245, as well as a similar interaction of the oxadiazole ring's nitrogen with the amino side chain of Asn241. A face-to-face π-π stacking was present between the quinoline ring and Phe300. The ligand's stability at the allosteric site was fostered by hydrophobic interactions with Lys155, Ser156, Phe157, Phe177, His239, Pro240, His245, Ala278, and Phe311 residues. More importantly, these interactions were anchored by a halogen bond interaction of the bromopentyl unit with Agr312, Tyr313, and Asn412, an interaction which conceivably accounts for the compound's superior αglucosidase inhibition. The biochemical data indicated compound 12k as the second most potent α-glucosidase inhibitor, and this was reflected in the ligand's docking score of −6.757. Interestingly, the imidazole ring of His245 flipped 180° to form a hydrogen bond interaction with the ligand's quinoline nitrogen, while the oxadiazole ring nitrogen and oxygen atoms formed hydrogen bond interactions with Arg439 and amino group of Arg312, respectively. Also, the triazole ring nitrogen formed a hydrogen bond with Agr312 even as the quinoline ring established a π-π stacking interaction with Phe157. Hydrophobic interactions with Phe157, Phe177, Thr215, Leu218, Hie245, Ala278, Phe300, and Arg312 residues stabilized the 12k-α-glucosidase complex. Furthermore, the compound 4k-α-glucosidase complex exhibited a lower docking score (−6.498) compared with 4i & 12k complexes, aligning with the ligands' IC50 values. Compound 4k's interaction profile consisted of hydrogen bond interactions of 8-HQ oxygen with Phe300, oxadiazole ring hydrogen bond donors spanning to the opposite ends of Arg312, one on the side chain amino group, and peptide bond NH. The aromatic rings of Phe157 and Phe300 aligned in plane to establish a π-π stacking interaction with the quinoline ring.
For compound 4l-α-glucosidase complex, the hydrogen bond interaction of the quinoline nitrogen was present with His239, while oxygen atoms of the ether linkage and oxadiazole ring interacted in a like manner with Lys155. The oxadiazole nitrogen also formed additional hydrogen bonds with Lys155 and Gly160, respectively. Moreover, the quinoline ring underwent a π-π stacking interaction with the His239 imidazole ring, while the oxadiazole ring experienced a π-cation stacking interaction with the protonated nitrogen of Lys155. The aromatic ring chlorine formed a halogen bond with Asn314, and complex stability was reinforced by hydrophobic interactions with Ly155, Phe157, Phe310, and Glu426. The biochemical data indicated compound 12k as the second most potent α-glucosidase inhibitor, and this was reflected in the ligand's docking score of −6.757. Interestingly, the imidazole ring of His245 flipped 180 • to form a hydrogen bond interaction with the ligand's quinoline nitrogen, while the oxadiazole ring nitrogen and oxygen atoms formed hydrogen bond interactions with Arg439 and amino group of Arg312, respectively. Also, the triazole ring nitrogen formed a hydrogen bond with Agr312 even as the quinoline ring established a π-π stacking interaction with Phe157. Hydrophobic interactions with Phe157, Phe177, Thr215, Leu218, Hie245, Ala278, Phe300, and Arg312 residues stabilized the 12k-α-glucosidase complex. Furthermore, the compound 4k-α-glucosidase complex exhibited a lower docking score (−6.498) compared with 4i & 12k complexes, aligning with the ligands' IC 50 values. Compound 4k's interaction profile consisted of hydrogen bond interactions of 8-HQ oxygen with Phe300, oxadiazole ring hydrogen bond donors spanning to the opposite ends of Arg312, one on the side chain amino group, and peptide bond NH. The aromatic rings of Phe157 and Phe300 aligned in plane to establish a π-π stacking interaction with the quinoline ring.
For compound 4l-α-glucosidase complex, the hydrogen bond interaction of the quinoline nitrogen was present with His239, while oxygen atoms of the ether linkage and oxadiazole ring interacted in a like manner with Lys155. The oxadiazole nitrogen also formed additional hydrogen bonds with Lys155 and Gly160, respectively. Moreover, the quinoline ring underwent a π-π stacking interaction with the His239 imidazole ring, while the oxadiazole ring experienced a π-cation stacking interaction with the protonated nitrogen of Lys155. The aromatic ring chlorine formed a halogen bond with Asn314, and complex stability was reinforced by hydrophobic interactions with Ly155, Phe157, Phe310, and Glu426.
The lower inhibitory potencies of compounds 4j and 12m (relative to 4b, 4i, 4k, 4l and 12k) were highlighted in their docking scores of −5.040 and −4.961, respectively. The interaction profiles of these two compounds were not so impressive as other examined ligands, corroborating the biochemical data. Although the oxadiazole nitrogen formed hydrogen bond interactions with His239 and Arg312 in compound 4j, and Ser281 and Tyr286 in compound 12m, the absence of the quinoline nitrogen's hydrogen bond interaction with allosteric site residues seen in other compounds presumably accounts for the low affinity that was recorded. These results jointly rationalize the compounds' behavior as non-competitive inhibitors i.e., they may reversibly bind at the allosteric site to trigger altered conformational dynamics at the enzyme's active site, consequently reducing catalytic activity without competing with the native substrate's binding.

Molecular Dynamics Simulations
Molecular dynamics (MD) simulations of the two most promising compounds 4i and 12k were performed over 200 ns trajectory, to further evaluate the ligands' stability at the allosteric site and the ligand-induced perturbation of protein conformation. Conformational stability was measured in terms of the root mean square deviations (RMSD) and root mean square fluctuations (RMSF) (Figure 7). RMSD indicates deviations of the protein-ligand complex from the reference structural conformation over the MD simulation trajectory, thus connoting the complex's fluctuations. Smaller fluctuations suggest a stable conformation of the protein-ligand complex, and vice versa.
The RMSD of the protein's C-α in the 4i complex experienced increased dynamics at the beginning of the MD simulation, highlighting the perturbation effect consequent to 4i binding at the allosteric site. RMSD for the protein and ligand reached equilibrium after 70 and 120 ns, respectively, with minimal fluctuations from the average values of 3.0 and 8.0 Å, thus indicating a stable complexation over the 200 ns simulation. In contrast, larger fluctuations persisted in the RMSD evolutions of the 12k-α-glucosidase model system over the 200 ns trajectory. This highlights the system's reduced stability and, consequently, the weaker α-glucosidase inhibition of 12k compared to 4i. Nonetheless, the 12k model system equilibrated after 100 ns with no significant differences in fluctuation, hence the protein's C-α and ligand's RMSD converged at 2.4 and 6.5 Å, respectively. More importantly, the RMSD convergence of the protein's C-α in both complexes was within the order of ≤3 Å showing no significantly large fluctuations, an indication of its excellent ability to maintain a stable residency, and ligands did not diffuse [39].
RMSF is significant for characterizing changes in protein residues' flexibility within the chain. Herein, the ligands' effect on the enzyme structure's flexibility for each residue fluctuation was studied. Figure 7b shows that compound 4i efficiently induced fluctuations in the protein's residues, as seen by its highest RMSF value at~3.8 Å with residues Ala144 and Lys233 fluctuating the most during the 200 ns simulation. Meanwhile, the highest RMSF in the compound 12k complex occurred at 4.5 Å with residues Lys233 and Ala289. It is also inferable from Figure 7b that compound 4i perturbed the receptor more than 12k did. The number of green vertical bars denoting the residues interacting with the ligand were higher in the compound 4i-α-glucosidase complex compared with 12k. This further establishes the superior docking score, affinity for the target's allosteric site, and consequently the α-glucosidase inhibitory potency of compound 4i relative to 12k.  RMSF is significant for characterizing changes in protein residues' flexibility within the chain. Herein, the ligands' effect on the enzyme structure's flexibility for each residue fluctuation was studied. Figure 7b shows that compound 4i efficiently induced fluctuations in the protein's residues, as seen by its highest RMSF value at ~3.8 Å with residues Ala144 and Lys233 fluctuating the most during the 200 ns simulation. Meanwhile, the highest RMSF in the compound 12k complex occurred at 4.5 Å with residues Lys233 and Ala289. It is also inferable from Figure 7b that compound 4i perturbed the receptor more than 12k did. The number of green vertical bars denoting the residues interacting with the ligand were higher in the compound 4i-α-glucosidase complex compared with 12k. This further establishes the superior docking score, affinity for the target's allosteric site, and consequently the α-glucosidase inhibitory potency of compound 4i relative to 12k.
the protein-ligand interactions were monitored over the simulation trajectory for contacts such as hydrogen bonds, hydrophobic, and water-mediated hydrogen bonding. Hydrogen bonds and hydrophobic contacts play significant roles at the binding site in ligands' affinity and stability, respectively, to elicit a biological response; hence their occurrence is vital in MD simulations [40]. The analysis (Figure 7c) showed that hydrophobic interactions with α-glucosidase dominated in compound 4i compared to 12k; thus, reaffirming compound 4i's improved α-glucosidase inhibition. The protein-ligand interactions were monitored over the simulation trajectory for contacts such as hydrogen bonds, hydrophobic, and water-mediated hydrogen bonding. Hydrogen bonds and hydrophobic contacts play significant roles at the binding site in ligands' affinity and stability, respectively, to elicit a biological response; hence their occurrence is vital in MD simulations [40]. The analysis (Figure 7c) showed that hydrophobic interactions with α-glucosidase dominated in compound 4i compared to 12k; thus, reaffirming compound 4i's improved α-glucosidase inhibition.

Conclusions
The antidiabetic potential of the quinoline pharmacophore via α-glucosidase inhibition and its antioxidant effects has been detailed in this contribution. Biological evaluations of the synthesized compounds (4a-t and 12a-t) suggested that the 1,3,4-oxadiazole core was more beneficial to the desired activity compared with 1,2,3-triazole. Compound 4i bearing a bromopentyl sidechain and compound 12k with an unsubstituted phenyl-1,2,3triazole pendant emerged as the most promising α-glucosidase inhibitors overall. Enzyme kinetics further revealed a non-competitive mode of inhibition, thus classifying the present compounds as allosteric α-glucosidase inhibitors. This behavior is valuable to reducing systemic glucose levels while evading the limitations associated with the competitive type of target inhibition. Additionally, molecular docking and molecular dynamics simulations corroborated the compounds' stability at the allosteric site, as evidenced by their strong interactions with the receptor. The results together illuminate the potential of the present structural template for developing new α-glucosidase inhibitors with improved efficacy for T2DM management.

Materials and Methods
All chemicals and reagents were purchased from Merck, Johannesburg, South Africa and were used without any purification. The reaction progress was monitored with thin layer chromatography (TLC) plates. Purification of the final products was achieved using column chromatography with silica gel (0.063-0.200 mm) at different gradients of EtOAchexane eluents. Melting points were determined with open-end capillary tubes in a Stuart melting point instrument (SMP-3) and were uncorrected. The 1 H, 13 C, and 2D-NMR spectra of all synthesized compounds were recorded on 400 and 600 MHz Bruker AvanceIII spectrometers. The chemical shifts were recorded in parts per million (ppm) with deuterated dimethyl sulfoxide-d 6 (δ H 2.50 and δ C 39.50 ppm) and chloroform-d (δ H 7.26 and δ C 77.00 ppm), wherein, tetramethylsilane (TMS) at δ H = 0 was used as an internal standard. The splitting pattern is abbreviated as singlet (s), doublet (d), triplet (t), quartet (q), multiplet (m), pentet (p), doublet of triplet (dt), doublet of doublet (dd), triplet of doublet (td) and doublet of quartet (dq), while coupling constants J are reported in Hertz (Hz). High-resolution mass spectra were recorded using a Water Micromass LCT Premier TOF-MS spectrometer. The general synthetic procedure for azides is outlined in the supporting information.

Synthesis of 5-[(Quinolin-8-yloxy)methyl]-1,3,4-oxadiazole-2-thiol (3)
Anhydrous potassium carbonate 138mmol (2.0 eq) was added to a solution of 8-HQ (68.88 mmol) in DMF 40mL in a round bottom flask and stirred at room temperature (r.t). After a few minutes, ethyl chloroacetate was added with continued stirring for 6 h. The reaction was stopped after TLC showed a consumed 8-HQ, then the mixture was poured into a slurry of crushed ice and the resulting precipitate compound 1 was collected via vacuum filtration as a beige solid at 88% yield. Compound 1 (14.00 g 60 mmol) was dissolved in absolute ethanol 80 mL, and 120 mmol, 2.0 eq of hydrazine hydrate was added slowly. The whole mixture was refluxed at 95 • C overnight. Thereafter, the solvent's volume was reduced and the reaction flask was cooled in an ice bath to precipitate 2-(quinolin-8-yloxy)acetohydrazide 2, which was filtered in vacuo while washing with cold diethyl ether to obtain 12.02 g of off-white solid (91% yield). Compound 2 (12.02 g, 55 mmol) was then poured into ethanolic potassium hydroxide and stirred at 80 • C for few a minutes before dropwise addition of carbon disulfide (221 mmol, 5eq). The mixture was refluxed at 80 • C for 12 h. Subsequently, the reaction mixture was evaporated to dryness and the crude product was treated with dilute HCl. The resulting precipitate was then poured into an ice slurry to yield a yellow solid. Recrystallization from ethanol afforded 5-((quinolin-8yloxy)methyl)-1,3,4-oxadiazole-2-thiol as pale yellow solid at 80% yield.

2-(Quinolin-8-yloxy)acetohydrazide (2)
Off-white solid; Chemical formula: C 11   To a solution of compound 3 (103.7 mg, 0.4 mmol) in DMF (3 mL), potassium carbonate (82.92 mg, 1.5 eq) was added and stirred until the reaction formed a paste in the round bottom flask. Then, 1.05 eq of appropriate benzyl bromides and aliphatic alkyl halides were added and stirring continued at r.t. for 1-2 h. Thereafter, the flask's contents were poured in a slurry of ice and extracted with ethyl acetate. The crude product was purified with column chromatography using EtOAc-hexane eluents to obtain quinoline-1,3,4-oxadiazole conjugates in moderate to quantitative yields.   A measure of 0.722 g (3.4 mmol 1 eq) of compound 3 was dissolved in a stirring mixture of DMF (5 mL) and pre-activated K 2 CO 3 (0.769 g, 2.0 eq). A sticky mass was formed after few minutes, then 0.36 mL of propargyl bromide (80% in toluene) in DMF (2 mL) was added and stirring continued at r.t for 4 h. On completion of the reaction as evidenced by TLC, the mixture was poured into a slurry of ice and stirred vigorously for 30 min. The resulting brown precipitate was filtered in vacuo to obtain 0.66 g compound 11 (79% yield).
Brown solid; Chemical formula: C 15  To a solution of alkyne intermediate 11 (0.15 g) in DCM (10 mL) in a round bottom flask, sodium ascorbate (22 mol%) and copper(II) sulphate pentahydrate (10 mol%) in water (10 mL) were added and stirred at r.t for few minutes. Then, the appropriate azides (1.1 eq) in DCM were added and the mixture was stirred for 2-4 h. After this time, TLC analysis showed the alkyne was consumed, then the reaction was diluted with water and filtered to remove residual salts. The resulting filtrate was extracted with DCM, washed with brine and dried over anhydrous sodium sulphate, filtered, and concentrated under reduced pressure. The crude product was purified with column chromatography in EtOAc-hexane eluents to obtain pure quinoline-1,3,4-oxadiazole-1,2,3-triazole hybrids 12a-t in excellent yields (43-91%).  13