Highly Accessible Computational Prediction and In Vivo/In Vitro Experimental Validation: Novel Synthetic Phenyl Ketone Derivatives as Promising Agents against NAFLD via Modulating Oxidoreductase Activity

Abstract

Nonalcoholic fatty liver disease (NAFLD) has reached epidemic proportions with no pharmacological treatment approved. Several highly accessible computational tools were employed to predict the activities of twelve novel compounds prior to actual chemical synthesis. We began our work by designing two or three hydroxyl groups appended to the phenyl ketone core, followed by prediction of drug-likeness and targets. Most predicted targets for each compound overlapped with NAFLD targets (≥80%). Enrichment analysis showed that these compounds might regulate oxidoreductase activity. Then, these compounds were synthesized and confirmed by IR, MS, ¹H, and ¹³C NMR. Their cell viability demonstrated that twelve compounds exhibited appreciable potencies against NAFLD (EC₅₀ values ≤ 13.5 μM). Furthermore, the most potent compound 5f effectively prevented NAFLD progression as evidenced by the change in histological features. 5f significantly reduced total cholesterol and triglyceride levels in vitro/in vivo, and the effects of 5f were significantly stronger than those of the control drug. The proteomic data showed that oxidoreductase activity was the most significantly enriched, and this finding was consistent with docking results. In summary, this validated presynthesis prediction approach was cost-saving and worthy of popularization. The novel synthetic phenyl ketone derivative 5f holds great therapeutic potential by modulating oxidoreductase activity to counter NAFLD.

Figure 1

Workflow of this study. The contents within the blue box are performed on computers. The contents within the red box are traditional and more realistic experiments. nBu: n-butyl; iBu: isobutyl; tBu: tert-butyl.

Figure 2

Computational prediction results of twelve compounds 5a-l. (a) Drug-likeness features of compounds 5a-l. (b) The percentage of predicted targets of each compound overlapped with targets of NAFLD. (c) The predicted targets on the petal diagram; each petal represents the unique targets of a compound, different colors represent different compounds, and the number in the middle represents 63 common targets of these twelve compounds. (d) Gene Ontology (GO) molecular function term of 63 common predicted targets.

Figure 3

Synthetic routes of target compounds 5a-l.

Figure 4

Effects of compound 5f in HepG2 cells and mice. (a) Cellular triglyceride (TG) accumulation. (b) The total cellular cholesterol (TC) level. (c) A schematic illustration of the animal experiment. (d) The weight growth of mice. (e) Serum TG accumulation. (f) Serum TC accumulation. (g) Representative images of hematoxylin and eosin- (H&E-) stained sectioned livers from the mice. Data are mean ± SD; compared with that in the control group, ^#p < 0.05 and ^##p < 0.01; compared with that in the model group, ^∗p < 0.05 and ^∗∗p < 0.01. ns: not significant; Ctrl: control; PA: palmitic acid; Sil: silybin; ND: normal diet; HFD: high-fat diet.

Figure 5

Proteomics results of compound 5f-treated HepG2 cells. (a) The top fifteen GO molecular function terms. (b) The relative protein level of twenty-five differentially expressed proteins (DEPs) which were clustered in the term “oxidoreductase activity.” DEPs are presented as official gene names and are arranged from left to right in order of increasing p value.

Figure 6

The results of molecular docking. (a) The root-mean-square deviation (RMSD) value of compound 5f or sil with nineteen DEPs clustered in the term “oxidoreductase activity.” (b) The docking score of compound 5f or sil with nineteen DEPs clustered in the term “oxidoreductase activity.” (c–e) The visualizations of 5f with NQO1. 5f is depicted in capped sticks representation, while NQO1 is shown as gold ribbon. The ligand surface is indicated in magenta. The interacting residues are shown in line representation. H-bond interactions are in yellow dotted lines.