Computational Design of Potent and Selective d-Peptide Agonists of the Glucagon-like Peptide-2 Receptor

Abstract

Here, we designed three d-GLP-2 agonists that activated the glucagon-like peptide-2 receptor (GLP-2R) cyclic adenosine monophosphate (cAMP) accumulation without stimulating the glucagon-like peptide-1 receptor (GLP-1R). All the d-GLP-2 agonists increased the protein kinase B phosphorylated (p-AKT) expression levels in a time- and concentration-dependent manner in vitro. The most effective d-GLP-2 analogue boosted the AKT phosphorylation 2.28 times more effectively compared to the native l-GLP-2. The enhancement in the p-AKT levels induced by the d-GLP-2 analogues could be explained by GLP-2R's more prolonged activation, given that the d-GLP-2 analogues induce a lower β-arrestin recruitment. The higher stability to protease degradation of our d-GLP-2 agonists helps us envision their potential applications in enhancing intestinal absorption and treating inflammatory bowel illness while lowering the high dosage required by the current treatments.

Figure 1

Design strategy of a novel d-peptide agonist of the glucagon like-2 receptor. (A) 3D structure of GLP-2 extracted from the complex with the glucagon like-2 receptor. The GLP-2 structure was divided into three fragments for running the query process: helix1 (H1-L14), helix2 (S7-A19), and helix3 (D15-I31). The dotted lines indicated the length of each helix. For clarity, only the hotspots selected in GLP-2 (H1, F6, D8, E9, L14, L17, D21, F22, and W25) were highlighted as licorice. (B) root mean square deviation (RMSD) profiles of the best output structures of each helix in the d-PDB database search. The structural superposition of the best d-retro-inverted match obtained for each helix was also shown. (C) Assembly of the d-retro-inverted GLP-2 analogue (d-GLP-2, cyan) by joining the three d-peptide matches obtained from the d-PDB database search. Structural superposition of d-GLP-2 over GLP-2 (light green). The hotspots and matching residues in GLP-2 and d-GLP-2, respectively, are highlighted as licorice.

Figure 2

Modeling the 3D structure of the GLP-2 receptor bound to d-GLP-2. (A) Structural superposition of d-GLP-2 (red cartoon) over the 3D structure of GLP-2 (yellow cartoon) bound to the GLP-2R (blue cartoon). The 3D structure of the GLP-2R in complex with the peptides was embedded in a POPC/PSM membrane with a composition of 1:1. The P atoms in the phospholipid polar heads are colored in cyan (POPC) and magenta (PSM). (B) RMSD of the heavy atoms of d-GLP-2 and GLP-2 bound to the GLP-2R along the MD simulations. (C) RMSF per residue of the heavy atoms of d-GLP-2 and GLP-2 bound to the GLP-2R along the MD simulations. The most representative cluster extracted from the MD simulations of the d-GLP-2 (red) and GLP-2 (yellow) bound to the GLP-2R (blue) were superimposed. (D) Zoomed image of the structural alignment at helix1. (E) Zoomed image of the structural alignment at helix2. (F) Zoomed image of the structural alignment at helix3. The hotspots (yellow) and matching (red) residues were highlighted as licorice. The nitrogen and oxygen atoms were colored blue and red, respectively, in the hotspots and matching residues.

Figure 3

Computational redesign of d-GLP-2. (A) Computational alanine scanning over the MD simulation of the GLP-2R + d-GLP-2 complex using the MM/GBSA method. The binding free energy differences were calculated as ΔΔG = ΔG_mutant – ΔG wild type, where ΔΔG > 0 indicates a favorable contribution to binding affinity. In contrast, ΔΔG < 0 shows an unfavorable contribution to the complex formation. (B) Zoomed image of the binding interaction of position E33 in d-GLP-2 with the GLP-2R. (C) Predicting the change in the binding affinity of d-GLP-2 of mutation E33A. (D) Predicting the change in the binding affinity of d-GLP-2 of mutation E33F. (E) Predicting the change in the binding affinity of d-GLP-2 of mutation E33Y. We used the CGI method combined with the dual-system single-box approach to calculate the binding free energy of the three mutations. (F) Zoomed image of the binding interaction of position H34 in d-GLP-2 modified with the NH2 group with the GLP-2R. (G) Zoomed image of the binding interaction of position H34 in d-GLP-2 modified with the NH–NH2 group with the GLP-2R.

Figure 4

CD measurements of the designed d-peptides and l-GLP-2 in solution. Peptide concentrations were 20 μM for l-GLP-2 and 150 μM d-GLP-2 in acetonitrile/PBS (1:2). Secondary structure determination was carried out using a Jasco J-720 spectro-polarimeter. Samples were read using a 0.1 cm cuvette path length with three accumulations per run and 50 nm/min scanning speed.

Figure 5

Experimental validation of the d-GLP-2 analogues. (A) Activity profile of l-GLP-2 and the design d-peptides over HEK293 cells stably expressing GLP-2R and CRE-luciferase. (B) Activity profile of l-GLP-2 and the design d-peptides over control HEK293 cells stably expressing GLP-1R and CRE-luciferase. (C) A schematic illustration of the β-arrestin-2 recruitment assay. The reporter cells co-expressed the GLP-2R tagged with ProLink (PK) and β-arrestin-2 tagged with enzyme acceptor (EA). Upon activation of the GLP-2R-PK, there is recruitment of β-arrestin-2-EA, which in turn led to the complementation of the two β-galactosidase enzyme fragments (EA and PK), thereby hydrolyzing the substrate to generate a chemiluminescent signal. (D) Dose-response curve for β-arrestin-2 recruitment of GLP-2R by l-GLP-2 and the design d-peptides. (E) Western blots showing the GFP and protein kinase B phosphorylated (p-AKT) expression levels induced by the d-GLP-2 analogues and the native l-GLP-2 at 10 μM. We measured the GFP expression at 36 h, while the p-AKT expression was evaluated at 3 h. We measured the β-actin expression in all the treatments as loading controls. (F) Quantification of the GFP expression for all the peptides relative to the l-GLP-2-treated cells after 36 h of incubation. (G). Quantification of the AKT phosphorylation induced for all the peptides relative to the l-GLP-2-treated cells after 3 h of incubation. In (FandG) *P < 0.05, **P < 0.01 versus l-GLP-2-treated cells; n.s., not significant.

Figure 6

The design d-GLP-2 analogues are resistant to DPP-IV and proteinase K degradation. (A) Sample gel images of the designed d-peptides and l-GLP-2 treated with DPP-IV over 150 min. Gels were stained with Coomassie brilliant blue dye. Band densitometries were calculated using ImageJ with background subtraction. (B) Quantification of the remaining peptides after DPP-IV treatment at 30 min intervals. Intensities of peptide bands were normalized to the intensity of the untreated peptide (T0). (C) Sample gel images of the designed d-peptides and l-GLP-2 treated with proteinase K over 150 min. Gels were stained with Coomassie brilliant blue dye. Bands densitometries were calculated using ImageJ with background subtraction. (D) Quantification of the remaining peptides after proteinase K treatment at 30 min intervals. Intensities of peptide bands were normalized to the intensity of the untreated peptide (T0).