Elucidation of Pharmacological Mechanism Underlying the Anti-Alzheimer's Disease Effects of Evodia rutaecarpa and Discovery of Novel Lead Molecules: An In Silico Study

Abstract

Alzheimer's disease (AD) is a brain disease with a peculiarity of multiformity and an insidious onset. Multiple-target drugs, especially Chinese traditional medicine, have achieved a measure of success in AD treatment. Evodia rutaecarpa (Juss.) Benth. (Wuzhuyu, WZY, i.e., E. rutaecarpa), a traditional Chinese herb, has been identified as an effective drug to cure migraines. To our surprise, our in silico study showed that rather than migraines, Alzheimer's disease was the primary disease to which the E. rutaecarpa active compounds were targeted. Correspondingly, a behavioral experiment showed that E. rutaecarpa extract could improve impairments in learning and memory in AD model mice. However, the mechanism underlying the way that E. rutaecarpa compounds target AD is still not clear. For this purpose, we employed methods of pharmacology networking and molecular docking to explore this mechanism. We found that E. rutaecarpa showed significant AD-targeting characteristics, and alkaloids of E. rutaecarpa played the main role in binding to the key nodes of AD. Our research detected that E. rutaecarpa affects the pathologic development of AD through the serotonergic synapse signaling pathway (SLC6A4), hormones (PTGS2, ESR1, AR), anti-neuroinflammation (SRC, TNF, NOS3), transcription regulation (NR3C1), and molecular chaperones (HSP90AA1), especially in the key nodes of PTGS2, AR, SLCA64, and SRC. Graveoline, 5-methoxy-N, N-dimethyltryptamine, dehydroevodiamine, and goshuyuamide II in E. rutaecarpa show stronger binding affinities to these key proteins than currently known preclinical and clinical drugs, showing a great potential to be developed as lead molecules for treating AD.

Keywords: Alzheimer’s disease; Evodia rutaecarpa; hormones; migraines; neurotransmitter inflammation; pharmacology network.

Figure 1

Workflow scheme. Our study was composed of three main steps. The first step consisted of compound screening and target prediction, then DO enrichment to focus on the targeted disease. The second step consisted of AD-related gene collection and overlapping with compound target genes, followed by a PPI network, GO and KEGG pathway enrichment, and C-T (overlapped) network construction, to find the core node proteins and pathways targeting AD. The third step consisted of molecular docking to further identify the underlying mechanism of WZY compounds in treating AD.

Figure 2

Compound-target network of 32 active compounds and bubble plot of DO enrichment analysis results of 229 target genes. (A) Compound–target network of 32 active compounds and target genes of WZY. The diamond nodes represent the compounds (pink for quinolone alkaloids, blue for indole alkaloids, yellow for other alkaloids, orange for fatty acids/esters, purple for sterols). The green octagon nodes with a red border represent the target genes. The nodes’ size changes from small to large as their degree value increases. (B) Bubble plot of top 20 DO enrichment analysis results of 229 WZY target genes, 1 × 10^−12.8 < Q value < 1 × 10^−4.3. The notable diseases are marked by red boxes. (C) Bubble plot of top 20 DisGeNET enrichment results of 229 WZY target genes, 1 × 10⁻⁵¹ < Q value < 1 × 10⁻²⁸.

Figure 3

Bubble plots of GO terms and bar plot of KEGG pathway enrichment analysis of the 229 target genes. The AD-related and notable terms and pathways are marked by red boxes. (A) Top 20 significant terms of biological processes’ enrichment result, 1 × 10^−42.3 < Q value < 1 × 10^−14.4. (B) Top 15 significant terms of cellular components’ enrichment result, 1 × 10^−17.2 < Q value< 1 × 10^−2.25. (C) Top 20 significant terms of molecular functions’ enrichment result, 1 × 10^−29.8 < Q value< 1 × 10^−5.3. (D) Top 20 significant terms of KEGG pathway enrichment result, 1 × 10^−73.8 < Q value < 1 × 10^−6.2.

Figure 4

Venn, DO, and PPI plots of 87 intersection genes. (A) Venn diagram of WZY active compounds overlapping with AD-related genes. (B) Bubble plot of top 20 significant terms of DO enrichment of 87 overlapping genes, 1 × 10^−27.3< Q value <1 × 10^−7.8, the AD-related and notable terms and pathways are marked by red boxes. (C) PPI network of 87 overlapping target genes. Clusters with different MCODE scores are represented by different colors, as the legend shows. The size of the node is larger as its degree value increases. The 9 nodes with the highest degree values are represented by a yellow border.

Figure 5

GO and KEGG pathway enrichment analyses results and compound–target network of 87 intersection genes. The AD-related and notable terms and pathways are marked by red boxes. (A) Top 20 significant terms of biological processes’ enrichment result, 1 × 10^−25.8 < Q value < 1 × 10^−10.3. (B) Top 9 significant terms of cellular components’ enrichment result, 1 × 10^−12.6 < Q value < 1 × 10^−2.21. (C) Top 19 significant terms of molecular functions’ enrichment result, 1 × 10⁻¹³ < Q value < 1 × 10^−2.36. (D) Top 20 significant terms of KEGG pathway enrichment result, 1 × 10^−18.7 < Q value < 1 × 10⁻³. (E) Compound–target network of 32 WZY active compounds targeting 87 intersection genes. The “V” nodes represent compounds (purple border for alkaloids). The round nodes represent target genes (pink border for 9 core AD-related genes). The size of the node is larger as its degree value increases. As the legend shows, the color of nodes changes from green to yellow to red along with their degree value.

Figure 6

Scheme of binding energy (kcal/mol) of molecular dockings of WZY active compounds binding to 9 key node proteins of AD. For each macromolecule, the compound with lowest binding affinity energy (strongest binding affinity) is marked with a purple-dotted square border. Dockings with lower binding affinity energy than the positive control are marked with purple “*”.

Figure 7

Molecular docking analysis of 9 core AD-related proteins bound to compounds with highest affinities (correspond to dockings with purple-dotted squares shown in Figure 6). (A) Sites of WZY25 (graveoline) binding to AR (PDB ID 5CJ6). (B) Sites of WZY14 (fordimine) binding to HSP90AA1 (PDB ID 1YC1). (C) Sites of WZY14 (fordimine) binding to ESR1 (PDB ID 6VIG). (D) Sites of WZY25 (graveoline) binding to ESR1 (PDB ID 6VIG). (E) Sites of WZY21 (goshuyuamide II) binding to NOS3 (PDB ID 7JRA). (F) Sites of WZY27 (dehydroevodiamine) binding to SRC (PDB ID 6E6E). (G) Sites of WZY27 (dehydroevodiamine) binding to PTGS2 (PDB ID 6E6E). (H) Sites of WZY16 (rutaecarpine) binding to SLC6A4 (PDB ID 6VRH). (I) Sites of WZY29 (rhetsinine) binding to NR3C1 (PDB ID 6DXK). (J) Sites of WZY17 (dihydrorutaecarpine) binding to TNF (PDB ID 6NH8).

Figure 8

Binding sites’ comparison of potential compounds versus positive control ligands. (A) Binding site comparison of WZY25 (graveoline) versus 51Y (positive control) in docking with AR (PDB ID 5CJ6). (B) Binding site comparison of WZY27 (dehydroevodiamine) versus mefenamic acid (positive control) in docking with PTGS2 (PDB ID 5IKR). (C) Binding site comparison of WZY26 (dehydroevodiamine) versus paroxetine (positive control) in docking with SLC6A4 (PDB ID 6VRH). (D) Binding site comparison of WZY26 (5-methoxy-N, N-dimethyltryptamine) versus HVY (positive control) in docking with SRC (PDB ID 5IKR).