3.12. Molecular Modeling Docking
The 3D crystal structure of the acetylcholinesterase complexed with acetylcholine (PDB: 2ACE) from
Torpedo californica exhibits that the substrate binding site of acetylcholinesterase is formed by 14α-helices, 14β-sheets, and numerous loops in a gorge shape [
28]. Once acetylcholine enters the substrate binding pocket by leaning its acetyl group toward the catalytic site (esteric site) and resides its trimethylamine group to the anionic site, it is surrounded by hydrophilic and hydrophobic residues including Trp 84, Gly 119, Glu 199, Ser 200, Phe 288, Phe 299, Glu 327, Phe 331, and His 440. Among these residues, Ser 200, Glu 327, and His 440 are viewed as the key residues at the esteric site to perform the hydrolytic reaction for acetylcholine when Ser 200 acts as the nucleophile. Furthermore, Trp 84 is considered as the essential residue at the anionic site. For the interactive binding mode of acetylcholine in the substrate binding pocket, the carbonyl group of acetylcholine is reduced to the hydroxyl group, which further interacts with the backbone of Gly 119 by acting as the H-bond acceptor. Apart from this H-bond interaction, there is no significant hydrophilic interaction between acetylcholine and the substrate binding pocket, especially the trimethylamine group of acetylcholine, which has been hypothesized to possibly make essential ionic interaction with the residues at the anionic site.
For the binding mode of galanthamine (
Figure 2a) in the substrate binding pocket of acetylcholinesterase, the crystal structure of the acetylcholinesterase complexed with (−)-galanthamine (PDB: 1W6R) from
Torpedo californica has been disclosed [
29]. Galanthamine binds to the substrate binding pocket by leaning its A, B, and C rings at a similar position as the acetyl group of acetylcholine, and locates its D ring at the anionic site. Since the A, B, and C rings of galanthamine are nearby Ser 200 and His 440, their three hydrophilic groups can form hydrophilic interactions with the esteric site, including (1) the 9-methoxy group on the A ring interacts with Ser 200 by performing as the H-bond acceptor, (2) the oxygen atom on the B ring also serves as the H-bond acceptor to interact with Ser 200, and (3) the 14-hydroxyl group on the C ring contacts with Glu 199 by serving as the H-bond donor. Additionally, the D ring of galanthamine, containing a tertiary amino group, which does not exhibit obvious ionic interaction, employs the methyl substituent on the amino group to make the nonclassical H-bond interaction with Asp 72. More importantly, galanthamine makes several essential π–π interactions in the substrate binding pocket, including (1) the A ring of galanthamine interacts with Phe 331, (2) the C ring of galanthamine contacts with Trp 84, and (3) the D ring of galanthamine interacts with Trp 84 and Phe 330.
To further study the interaction between catechin (
1) (
Figure 2b) (or gallic acid (
2)) and acetylcholinesterase, and try to interpret how
1 (or
2) might exert its antagonistic effect, The crystal structure of the acetylcholinesterase from
Electrophorus electricus (PDB: 1C2B) was also used in this study [
30]. The crystal structures of acetylcholinesterase from
Electrophorus electricus and
Torpedo californica share high sequence homology and have very similar conformation. In the substrate binding site for both structures, the key residues, including Ser 203 (Ser 200 in
Torpedo californica), His 447 (His 440 in
Torpedo californica), Glu 334 (Glu 327 in
Torpedo californica), Trp 86 (Trp 84 in
Torpedo californica), Tyr 133 (Tyr 130 in
Torpedo californica), Phe 338 (Phe 331 in
Torpedo californica), and Tyr 449 (Tyr 442 in
Torpedo californica), are almost identical. For the binding model of
2,
2 resided in the substrate binding pocket by leaning its three hydroxyl substituents toward the esteric site and locating its benzoic acid moiety at the anionic site. Compound
2 made two H-bond interactions comprising (1) the 3-hydroxyl group docked with the backbone of Gly 120 by serving as the H-bond donor, and (2) the carboxylate group made H-bond contact with Tyr 124 and Ser 125 by performing as the H-bond acceptor. Additionally,
2 made the essential π–π interaction with Trp 86 at the anionic site.
The docking model of catechin (
1) exhibited that
1 resided at a similar position as galanthamine, as shown in
Figure 3. In the substrate binding site, the B ring of
1 as the B ring of galanthamine leaned toward Ser 203 at the terminus of the α-helice shown by the purple color in
Figure 3, and the A ring of
1 as the C ring of galanthamine located nearby Trp 86. However, the C ring of
1 resided at a different position than the D ring of galanthamine. The C ring of
1 stayed between Tyr 124 and Phe 338, but the D ring of galanthamine was located between Trp 84 (Trp 86 in
Electrophorus electricus) and Phe 331 (Phe 338 in
Electrophorus electricus). Once compound
1 entered the substrate binding pocket, it made significant hydrophilic and hydrophobic interactions, including (1) the 3-hydroxyl group acted as the H-bond donator to make contact with Ser 203 and also served as the H-bond acceptor to interact with the backbone of Gly122; (2) the A ring of
1 contacted with Trp 86 and Tyr 337 by π–π interaction; and (3) the C ring of
1 interacted with Tyr 124, Phe 338, and Tyr 341 by π–π interaction.
The substrate binding pocket of acetylcholinesterase mainly includes two important sites, esteric and anionic sites. The esteric site contains some significant hydrophilic residues such as Glu 202 (Glu 199 in
Torpedo californica), Ser 203 (Ser 200 in
Torpedo californica), His 447 (His 440 in
Torpedo californica), and Glu 334 (Glu 327 in
Torpedo californica), and the potential acetylcholinesterase antagonist should have the structural moiety to make contact with these residues for good inhibition activity. For example, in galanthamine derivatives, whose 9-methoxy group is replaced by a phenoxy or 14-hydroxyl group is replaced by a carbonyl group, their antagonistic effect is largely decreased [
31]. For catechin (
1), galanthamine, and gallic acid (
2),
1 and galanthamine make similar H-bond interactions, but
2 only makes one H-bond interaction with the backbone of the unimportant residue at the esteric site. Based on these results, the antagonistic effect of
2 should be lower than that of
1 or galanthamine. On the contrary, the anionic site of acetylcholinesterase mainly contains hydrophobic residues such as Trp 86 (Trp 84 in
Torpedo californica), Tyr 133 (Tyr 130 in
Torpedo californica), Tyr 341 (Tyr 334 in
Torpedo californica), Phe 338 (Phe 331 in
Torpedo californica) and Tyr 449 (Tyr 442 in
Torpedo californica), so the potential acetylcholinesterase antagonist should have structural moieties to interact with these residues for good inhibition activity. Additionally, the tertiary amino moiety of the potential acetylcholinesterase antagonist located at the anionic site also plays an important role in the antiacetylcholinesterase activity, although it does not exhibit obvious ionic or hydrophilic interaction in the substrate binding pocket. The QSAR result shows that the galanthamine derivative containing a C(3) = N(4) double bond exhibits a better antiacetylcholinesterase effect than that having a C(3)–N(4) single bond. Since the anionic site of acetylcholinesterase contains key aromatic residues, the tertiary amino moiety might interact with the anionic site by the aromatic hydrophobic interaction rather than the ionic interaction. In particular, Trp 86 (Trp 84 in
Torpedo californica) is the key residue at the anionic site, and it is frequently shown to exhibit a π–π interaction with the acetylcholinesterase antagonist. For example, the crystal structure of the acetylcholinesterase complexed with huperzine A (PDB: 1VOT) from
Torpedo californica indicates that huperzine A does not make an H-bond interaction with Ser 200 or Glu 199 at the esteric site, and the pyridine moiety of huperzine A makes a strong π–π interaction with Trp 84 [
28]. Furthermore, for the binding modes of
1 and galanthamine at the anionic site, the key residue Trp 86 (Trp 84 in
Torpedo californica), only contacts with the A ring of
1, but Trp 84 makes three π–π interactions with the C and D rings of galanthamine. According to the evidence mentioned above, it is highly exhibited that the difference in the binding modes of
2,
1, and galanthamine at the anionic site might lead to their distinct antiacetylcholinesterase potency.
In addition, the interaction between euscaphic acid (
5) (
Figure 4) and acetylcholinesterase was also evaluated. The crystal structure of the acetylcholinesterase from
Electrophorus electricus (PDB: 1C2B) was employed in this study. For the binding model of
5 (
Figure 5), the A ring of
5 stayed between Tyr 124 and Phe 338, a similar position to the C ring of
1. However, the A ring of
5 resided at a different position from the D ring of galanthamine. The D ring of galanthamine was located between Trp 84 (Trp 86 in
Electrophorus electricus) and Phe 331 (Phe 338 in
Electrophorus electricus).
5 made two H-bond interactions comprising (1) the 2-hydroxyl group on the A ring of
5 interacts with Phe 295 by performing as the H-bond acceptor, and (2) the 19-hydroxyl group on the E ring of
5 interacts with Trp 286 by performing as the H-bond acceptor. Additionally, the A ring of
5 interacted with Tyr 124, Tyr 341, and Phe 297 by π–alkyl interaction.
The lowest binding energy of each ligand was regarded as the optimal conformation. In this study, chlorogenic acid and galanthamine were employed as positive control. The binding energies of compounds
5,
1,
2,
3,
7,
4, and
6 were −8.7, −8.5, −8.3, −8.1, −8.1, −7.8, and −7.6 kcal/mol, respectively (
Table 7). Compared with chlorogenic acid, the binding energies of compounds
1–
3,
5, and
7 were lower than −8.0 kcal/mol. This shows that
1–
3,
5, and
7 can dock into the pocket of the crystal structure of the acetylcholinesterase from
Electrophorus electricus more effectively than that of chlorogenic acid.
To further study the interaction between compounds 3–5 and α-glucosidase. The crystal structure of the isomaltase from Saccharomyces cerevisiae (PDB: 3A4A) was also employed in this study. Compounds 3–5 showed potent α-glucosidase inhibitory activity. Therefore, the interaction between 3–5 and α-glucosidase was evaluated by molecular model docking.
In this research, acarbose was employed as a positive control. The binding energies of compounds
3,
4,
5,
6,
1,
7, and
2 were −8.5, −8.0, −7.8, −7.8, −7.6, −7.5, and −5.1 kcal/mol, respectively (
Table 8). Compared with the positive control, the binding energies of compounds
1 and
3–
7 were less than −5.3 kcal/mol. This shows that compounds
1 and
3–
7 could dock into the pocket of the crystal structure of isomaltase from
Saccharomyces cerevisiae more effectively than that of acarbose.
As shown in
Figure 6, compound
3 was bound with Gln 353, Gln 277, and Asp 69 through conventional hydrogen bonds, while other interactions (π–π T-shaped, π–alkyl, and π–anion) were also observed with Phe 303, Tyr 72, Val 216, Asp 352, Glu 277, and Glu 411. These permitted
3 and the enzyme to create a stable complex.
For compound
4 (
Figure 7), binding to Glu 277 was via conventional hydrogen bonds, while other interactions (π–π T-shaped, π–alkyl, and π–cation) were also observed with Tyr 158, Arg315, and Arg 442.
Finally, compound
5 was bound with Arg 315 and Leu 313 via conventional hydrogen bonds, while other interactions (alkyl and π–alkyl) were also perceived with Arg 442, Phe 303, and Lys 156 (
Figure 8).
As a positive control, acarbose binds to Arg 442, Asp 69, Asp 215, Asp 352, Asn 415, and Gln 353 via conventional hydrogen bonds. Glu 277 forms a carbon–hydrogen bond with acarbose. Furthermore, two hydroxyl groups provide unfavorable donor–donor interactions with Arg 213 and Gln 279 residues (
Figure 9).
On the basis of our data, the docking binding energies of compounds 1 and 3–5 are lower than that of acarbose, suggesting that they have better binding capability. In our research, the active ingredients 1 and 3–5 not alone exhibited α-glucosidase inhibitory activity but, likewise, had stronger binding potentiality with the active sites of α-glucosidase from Saccharomyces cerevisiae. This suggests that these components can be worthy of further research as natural anti-α-glucosidase agents.