New Molecular Insights into the Inhibition of Dipeptidyl Peptidase-4 by Natural Cyclic Peptide Oxytocin

Abstract

Protease inhibition has led to treating many diseases and has been successful in producing many commercial drugs by pharmaceutical companies. Among many proteases, serine protease has been attractive in treating metabolic disorder diabetes mellitus (DM). Gliptins have been proven to inhibit dipeptidyl peptidase-4 (DPP4), a serine protease, and are an emerging therapeutic drug target to reduce blood glucose levels, but until now there is no natural cyclic peptide proven to inhibit serine protease DPP4. This study demonstrates the potential mechanism of natural cyclic peptide oxytocin (OXT) as a DPP4 inhibitor. To achieve this, initially, activity atlas and field-based models of DPP4 inhibitors were utilized to predict the possible features of positive and negative electrostatic, hydrophobic, and activity shapes of DPP4 inhibition. Oxytocin binding mode, flexibility, and interacting residues were studied using molecular docking simulations studies. 3D-RISM calculations studies revealed that the stability of water molecules at the binding site are favorable. Finally, an experimental study using fluorescence assay revealed OXT inhibits DPP4 in a concentration-dependent manner in a significant way (p < 0.05) and possess IC₅₀ of 110.7 nM. These new findings significantly expand the pharmaceutical application of cyclic peptides, and in specific OXT, and implicate further optimization of OXT inhibition capacity to understand the effect of DPP4 inhibition. This work highlights the development of natural cyclic peptides as future therapeutic peptides to reduce glucose levels and treat diabetes mellitus.

Keywords: cyclic peptides; diabetes treatment; dipeptidyl peptidase-4 inhibition; oxytocin; peptide therapeutics.

Figure 1

Three-dimensional structure of oxytocin visualized using Forge visualization software.

Figure 2

Activity atlas model describing the features responsible for DPP4 inhibition (left side) and features predicted for natural cyclic peptide oxytocin (right side). Forge tool was utilized to apply, analyze, and visualize features predicted by the activity atlas model. The generated model provides average electrostatics of positive actives in red color and negative electrostatics in blue color (a); an average of hydrophobicity (b); an average of actives (c); and activity cliff summary of favorable and unfavorable shapes (d).

Figure 3

Top scoring binding pose views of oxytocin at the active site of DPP4 (a); and two-dimensional study of interacting residues (b), involved at the binding pose of oxytocin.

Figure 4

The stability of water molecules at the oxytocin bound to DPP4 protein calculated using the three-dimensional reference interaction model (3D-RISM). Free energy (ΔG) is colored in green for favorable water molecules and in red color for the unfavorable water molecules, calculated for the oxygen isodensity surface ρ = 5. Cresset visualization software was used to visualize the molecular insights of water molecules’ stability at the complex.

Figure 5

Relative percent inhibition of the dipeptidyl peptidase-4 (DPP4) enzyme by natural cyclic peptide oxytocin at five different concentrations. The two-tailed unpaired t-test was utilized to calculate the significant difference among different concentrations. Data represent mean ± S.D.

Figure 6

Comparison of features predicted by the activity atlas model on clinically used inhibitor and natural cyclic peptide oxytocin. The three-dimensional structure of oxytocin and sitagliptin (a); average activity summary of positive is shown in red color and negative electrostatics is shown in blue color (b); average shape of hydrophobics active (c); and average shape of actives (d) calculated using activity atlas model of DPP4 inhibitors. Forge visualization tool is utilized to understand the features.