Computer-aided repositioning and functional in vitro assessment of novel PAD4 inhibitors

Abstract

Peptidyl arginine deiminase 4 (PAD4) is a protein that catalyzes both normal and abnormal citrullination of interacting protein partners, affecting gene regulation and being associated with diseases such as Alzheimer's, cancer, and rheumatoid arthritis (RA). As a result, PAD4 has emerged as a potential therapeutic target; however, no inhibitors have been approved to date. In this study, the REFRAME and ZINC15 drug databases were virtually screened. The approach used molecular docking and dynamics simulation techniques to identify compounds with high-predicted binding affinity to PAD4 (PDB ID: 4DKT). Selected hits from this virtual screening underwent in vitro assays using fixed concentrations derived from the docking score to evaluate their ability to inhibit PAD4 activity, and their effect on neutrophil extracellular trap (NET) release was assessed using an ex vivo human neutrophil model. Computational analyses identified amodiaquine, folic acid, and pyroxamide as stable PAD4 binders. In vitro inhibition assays revealed that amodiaquine (5.0 μM to 1.0 nM) and pyroxamide (0.1 μM) were more potent inhibitors than the reference PAD4 inhibitor BBCla (8.8 μM), while folic acid showed a non-significant trend toward inhibition. Cytotoxicity assays confirmed that all compounds were non-toxic at the tested concentrations, except for amodiaquine at 50 μM. NETosis assays demonstrated that the three selected compounds altered chromatin decondensation and cellular morphology similarly to BBCla, although not uniformly across all cells. Overall, amodiaquine, folic acid, and pyroxamide were identified as PAD4 inhibitors through combined virtual and experimental approaches, supporting their potential as therapeutic candidates for PAD4-related diseases and warranting further investigation.

Fig. 1. Clustering of selected compounds. Newick tree illustrates drug–receptor interactions and the grouping of compounds into two main groups.

Fig. 2. Interaction of selected compounds on the active site. (a) TDFA compound, (b) folic acid, (c) amodiaquine, and (d) pyroxamide. The dotted lines represent the hydrogen bonds, and in purple hydrophobic interactions are shown.

Fig. 3. Molecular dynamic analysis. (a) RMSD as a function of time, showing the structural stability of the protein (in gray) in the presence of the three drugs: folic acid (pink), amodiaquine (blue) and pyroxamide (purple). (b) RMSF of the protein under the same conditions, with local residue fluctuations plotted as a function of time. The vertical lines on the graph indicate the position of the active site amino acids. (c–e) Hydrogen bond formation during molecular dynamics simulation.

Fig. 4. Evaluation of PAD4 activity. The compounds (a) folic acid, (b) amodiaquine and (c) pyroxamide are shown. The enzymatic activity was measured based on the changes of absorbance at each of the tested concentrations of the compounds, as well as the inhibition control BBCla (8.8 μM). (d) PAD4 activity in the presence of the three compounds at their most effective concentrations. PAD4 activity is expressed as a percentage relative to the untreated control (100%). The results for folic acid and pyroxamide are presented as the median and interquartile range, and the Kruskal–Wallis statistical test was used. Dunn's multiple comparison test was used as a post hoc test to identify pairwise differences between groups. For amodiaquine, the mean and standard deviations are shown, and the one-way ANOVA statistical test was used. Tukey's multiple comparison test was used as a post hoc test to identify pairwise differences between groups.

Fig. 5. Viability of polymorphonuclear cells. The effect of the compounds was evaluated against untreated human blood cells obtained from a healthy donor (HD). Results are shown as the mean percentage relative to the negative control in three independent experiments. Histograms (a), (d), and (g) represent cells stimulated with folic acid; (b), (e), and (h) show cells in the presence of amodiaquine; and histograms (c), (f), and (i) correspond to cells stimulated with pyroxamide. Bar graphs (j) and (k) show the mean and standard deviation of cell viability for folic acid, amodiaquine, and pyroxamide, respectively. One-way ANOVA statistical test was used. For pyroxamide (l), the median and interquartile range are presented and a Kruskal–Wallis test and Dunn's multiple comparison test was used as a post hoc analysis to identify pairwise differences between groups.

Fig. 6. Evaluation of PAD4 inhibition in the ex vivo NETosis process. (a) Unstimulated cells (US cells). All other cells were stimulated with (b) A23187 as the inhibition control and (c) BBCla 8.8 μM and the potential compounds (d) folic acid 1.0 μM, (e) amodiaquine 0.1 μM and (f) pyroxamide 0.1 μM. (g) Nuclear area and (h) perimeter measurements are shown. Unstimulated cells are shown in gray, while cells treated with A23187 are shown in green. Folic acid (pink), amodiaquine (blue), and pyroxamide (purple) are shown for different concentrations. Arrows indicate examples of cells with an unaffected nucleus. Results are expressed as the median and interquartile range, and statistical analysis was performed using the Kruskal–Wallis test. Dunn's multiple comparison test was used as a post hoc test to identify pairwise differences between groups.