3.1 Structural analysis of native GLP-1 peptide
According to the literature study (Adelhorst K et al. 1994, Hareter A et al. 1997), His7, Gly10, Phe12, Thr13, Asp15, Phe28, and Ile29 are important for receptor binding, while His7, Gly10, Asp15, and Phe28 are important for receptor activation. These findings suggest that GLP-1's N-terminal portion is involved in receptor activation and engagement, whereas the C-terminal area is only involved in receptor binding (Fig. 2). As a result, it was more reasonable to construct short peptides by selecting critical N-terminal residues. We chose nine residues from His7 to Asp15, and five of the nine residues are crucial for receptor activation and binding.
3.3 Protein peptide docking
To confirm the active pose for the suggested short peptides, the wild-type short peptide was docked on 6X18 and then overlaid on a co-crystallized GLP-1 native peptide. Furthermore, the docking analysis of the wild-type short peptide provided multiple reference points that aided in the selection of the best-mutated peptides from a total of 171 generated short peptides. The PIPER pose energy and PIPER pose score for the wild-type peptide was − 588.378 kcal/mol and − 170.024 kcal/mol respectively. It was observed that His7 formed H-bond with Gln234, interacted with Arg310 via water mediated H-bond (HOH103) and formed π- π stacking with Trp306. Similarly, all the interactions ae shown in Fig. 5. Tyr 152, Arg 190, and Tyr 241 have been marked to be important in activating the GLP-1 receptor at the peptide binding site and stabilizing the peptide inside the pocket by keeping the active receptor conformation (Lie et al. 2018, Wootten D et al. 2016, Wootten D et al. 2016). The peptides for the following investigation were chosen based on criteria such as a) docked pose, b) important interactions, and PIPER pose score. The following conclusions were drawn after a thorough examination of individual peptide residues.
His7
Histidine is a polar residue with a cationic side chain, and Hareter et al. 1997 has discussed the importance of the imidazole moiety in receptor activation and binding. Several intriguing observations were made when His 7 was replaced with 19 different amino acids. The ‘oxygen atom’ of the carboxyl group for Cys7, Gly7, Asn7, Ser7, and Val7 interacted with Tyr152 via H-bond, 'oxyanion' and ‘oxygen' atoms of Asp7 formed salt-bridge and H-bond with Arg190, Benzene ring of Phe7 interacted with Arg190 via -cation interaction, carboxyl group ‘oxygen’ atom of Ile7, Gln7 interacted with Arg190 via the H-bond, NH3 + group of Lys7 interacted with Phe369 and Ala368 through π-cation and H-bond interaction respectively, guanidine group of Arg7 attracted a couple of interactions that include the H-bond and salt bridge with Glu364 and Glu387, imidazole ring, ‘oxygen’ atom of the carboxyl group, and NH group of Trp7 interacted with Tyr148, Arg190, and Glu387 via π-π and H-bond interactions respectively, benzene ring of Tyr7 interacted with Glu387 through π-cation interaction and the hydroxyl group of Tyr7 formed H-bond with Val370. All of these findings suggested that the aromatic ring, carboxyl group, and NH group are critical for the receptor's interaction with important amino acids. As a result, thirteen of the nineteen amino acids interacted with receptor residues, emphasizing the importance of the peptide's initial position and its potential to attract multiple connections. Because His7 contains all of these crucial characteristics, mimicking this amino acid group with others is challenging. Among all the nineteen peptides, Trp7 substituted peptide showed the highest PIPER pose energy of -669.755 kcal/mol and PIPER pose score of -224.9 kcal/mol. Therefore, interaction analysis, as well as the pose analysis and PIPER pose scores, indicated that Phe7, Trp7, Tyr7, Asp7, and Arg7 may be substituted for His7. Biological assays, on the other hand, will be necessary to back up the theory. (Supplementary Fig. 1 (a-s) and Supplementary Table 1). Figure 6 shows the 2D interaction diagram for “W7AEGTFTSD” peptide docked on glp-1 receptor.
Ala8
Alanine is an aliphatic amino acid with a non-polar side chain. In the wild-type peptide, this amino acid had no interaction. However, when we replaced Ala with other amino acids, we found that 'oxygen' atom of Asp8's carboxyl group showed H-bond interaction with Gln234, amine group of Gln8 formed H-bond with Tyr152, guanidine group of Arg8 interacted with Leu388 by forming H-bond, ‘oxyanion’ and ‘oxygen’ atoms of Glu8 showed H-bond and salt bridge formation with Tyr152 and Arg190 respectively, NH3+ group of Lys8 interacted with Glu364, Glu 387, and Thr391 via H-bond and salt bridge with Glu364, Hydroxyl group of Ser8 and Thr8 interacted with Thr298 and Gln234 respectively via H-bond formation, benzene ring of Tyr8 interacted with Trp284 through π- π interaction. Eight of the nineteen amino acids were shown to be involved in receptor interactions. Tyr8 contains an aromatic ring and the polar side chain showed the highest PIPER pose energy of -640.496 kcal/mol and PIPER pose score of -170.024. Gly8 did not demonstrate interaction with the receptor, but the peptide interacted with the receptor through all important residues such as Arg190 and Tyr152. In summary, Asp8, Glu8, Ser8, Lys8, and Tyr8 demonstrated improved contacts, poses, and PIPER scores, suggesting that these changes might be employed to test their activity in biological experiments. (Supplementary Fig. 2 (a-s) and Supplementary Table 1). Figure 7 shows the 2D interaction diagram for “HY8EGTFTSD” peptide docked on glp-1 receptor.
Glu9
Glutamate is a polar residue with an anionic side chain and in the wild-type peptide, it attracts H-bond and salt bridge interaction with Lys383 of the receptor. The substitution of Glu9 with other residues showed H- bond interaction of Ala9 backbone NH group with Glu387, guanidine group Arg9 interacting with Ala368 via h-bond, H-bond formation by SH group of Cys9 interacting with Glu387, ‘oxygen’ atom of the carbonyl group of Gln9, Asn9 interacts with Arg190 via H-bond, NH3+ group of Lys9 attracted interactions like H-bond and salt bridge with Glu364 and H-bond with Thr391, main-chain NH group of Ser9 interacted with Glu387 through H-bond, Hydroxyl group of Thr9 interacted with Thr391 through H-bond. ‘oxygen’ atom of the carboxyl group of Asp9 interacted with Tyr152 and Arg190 via H-bond and salt bridge formation. In one case Tyr9 did not participate in interaction but the peptide obtained all the essential interactions required for receptor binding and showed the highest PIPER pose energy and score of -663.079 kcal/mol and − 267.396 kcal/mol respectively. Here, ten out of nineteen residues interacted with the receptor but Ala9, Asp9, and Tyr9 might be better for potentially replacing polar side amino acid Glu9. (Supplementary Fig. 3 (a-s) and Supplementary Table 1). Figure 8 shows the 2D interaction diagram for “HAY9GTFTSD” peptide docked on glp-1 receptor.
Gly10
Glycine contains an aliphatic side chain and Gly10 did not show any interaction for the wild-type peptide. oxyanion’ atom of Glu10 formed π-cation interaction with Lys383, Asn10 amine group formed H-bond with Glu387, guanidine group of Arg10 interacted with Tyr152 via H-bond, ‘oxyanion’ and ‘oxygen’ atoms of Asp10 formed salt-bridge and H-bond with Lys197. SH group of Cys10 formed H-bond with Glu387, imidazole ring of His10 formed π- π interaction with Trp306, the amine group of Gln10 formed H-bond with Glu387, the hydroxyl group of Thr10 and Ser10 formed H-bond with Asp372 and Glu387 respectively, benzene ring and imidazole ring of Trp10 formed π- π interaction and H-bond with Trp306 and Trp297, the hydroxyl group of Tyr10 interacted with Arg 190 by forming H-bond. Therefore, eleven out of nineteen residues interacted in place of Gly10. As an outliner, Trp10 and Leu10 itself did not interact with the receptor but the peptide itself nicely interacted with the vital residues. Based on all the vital interactions, poses, and PIPER scores; peptides with Arg10, His10, Tyr10 can potentially replace Gly10. (Supplementary Fig. 4 (a-s) and Supplementary Table 1). Figure 9 shows the 2D interaction diagram for “HAER10TFTSD” peptide docked on glp-1 receptor.
Thr11,13
Threonine is a polar residue with a neutral side chain. The substitution study of Thr11 gave us the following observations. The backbone carbonyl group of Asp11 interacted with Lys197 via H-bond, the carboxyl group oxygen atom of Asn11 formed H-bond interaction with Arg380, the amine group of Gln11 interacted with Asp372 through H-bond, the backbone carbonyl group of Cys11 formed H-bond with Lys197, ‘oxyanion’ and ‘oxygen’ atoms of Glu11 showed H-bond and salt bridge formation with Arg380 and Lys383, imidazole ring and N atom of His11 showed π-π interaction and H-bond with Tyr148 and Tyr152 respectively. It was seen that out of nineteen residues only six made interaction with the receptor. (Supplementary Fig. 5) Similarly, the second Threonine of the peptide was substituted by other remaining amino acids but this time the position changed to 13 and the following were the observations. The guanidine group of Arg13 interacted with Gln140 via H-bond, Asn13 carbonyl group formed H-bond with Arg380, ‘oxyanion’ atom of Asp13 formed a salt bridge with Arg380, the ‘oxyanion’ and ‘oxygen’ atoms of Glu13 showed H-bond and salt bridge formation with Arg380, the benzene ring of Trp13 formed π-cation interaction with Lys197. Hence, Glu11, Gln11, and His11 could potentially replace Thr11 whereas, Asn13 and Asp13 could possibly mimic Thr13. Also, it was found that Gly13, Ala13, and Gln13 did not interact on their own but did bind to the receptor in a fashion that they acquired interactions with all the vital residues like Tyr152, Arg190, and Tyr241. (Supplementary Fig. 5 (a-s) ,7 (a-p) and Supplementary Table 1). Figure 10 shows the 2D interaction diagram for “HAEGQ11FTSD” peptide docked on glp-1 receptor.
Phe12
Phenylalanine is a polar amino acid with an aromatic side chain. The Guanidine group of Arg12 interacts to Tyr148 by forming H-bond, the SH group of Cys12 interacts with Thr298 via H-bond, NH3+ group of Lys12 formed H-bond and salt bridge with Tyr148 and Asp198 respectively, the hydroxyl group of Ser12 and Thr12 interacted with Thr298 and Trp306 through H-bond respectively. This meant that only five residues were successful in interacting with the receptor. Residues like Arg12, Cys12, Lys12, and Thr12 can be used for replacement. (Supplementary Fig. 6 (a-s) and Supplementary Table 1). Figure 11 shows the 2D interaction diagram for “HAEGTR12TSD” peptide docked on glp-1 receptor.
Ser14
Serine is a polar residue with an uncharged side chain containing a hydroxyl group. The guanidine group of Arg14 interacted with Glu292 by forming a salt-bridge, ‘oxyanion’ and ‘oxygen’ atoms of Asp14 formed a salt-bridge and H-bond with Arg380, N atoms in imidazole the ring of His14 formed two H-bond with Asp372 and Lys 383, benzene ring of Trp14 formed π cation interaction with Arg376, the hydroxyl group of Tyr14 interacted with Arg190 via H-bond formation. Five residues play role in interacting with the receptor, whereas, residues like Arg14 and Asp14 can only be used to replace Ser14. (Supplementary Fig. 8 (a-s) and Supplementary Table 1). Figure 12 shows the 2D interaction diagram for “HAEGTFTD14D” peptide docked on glp-1 receptor.
Asp15
Aspartic acid is a polar residue with an anionic side chain. (a) Guanidine group of Arg15 interacted with Tyr205 and formed π cation interaction, (b) Carbonyl ‘oxygen’ atom and NH2 group of Asn15 formed H-bond with Lys197 and Ser193 respectively, (c) The carbonyl group of Gln15 formed H-bond with Arg380, (d) The ‘oxyanion’ and ‘oxygen’ atoms of Glu15 formed a salt-bridge and H-bond with Arg190 and H-bond with Tyr152, (e) The hydroxyl group of Tyr15 interacted with Tyr152 via H-bond, (f) The benzimidazole ring of Trp15 formed two π cation interactions with Arg190 and also made π-π stacking with Tyr152. Therefore, Arg15, Asn15, Tyr15, and Trp15 can be further studied. (Supplementary Fig. 9 (a-s) and Supplementary Table 1). Figure 13 shows the 2D interaction diagram for “HAEGTFTSW15” peptide docked on glp-1 receptor.
If we summarise this docking study, it is found that many residues can potentially mimic the wild-type residue provided they have similar properties to attract the vital interactions. Below we have shown some of the residues that could potentially mimic the wild-type residues and maintain the interactions nearby. (See Fig. 14) The docking poses of 171 designed peptides was thoroughly analyzed and based on peptide poses, scores, and interactions gave us 37 peptides that can potentially be used for further studies. Out of 37 peptides, five were outliers as they did not show interaction through the mutated residue but as a peptide, they seemed to be quite promising by attracting numerous vital interactions with the receptor. The wild-type 9 residue short peptide contains three non-polar residues namely Ala8, Gly10, and Phe12, and six polar residues His7, Glu9, Thr11, Thr12, Ser14, Asp15. It was observed that positions 7, 9, 12, 15 attracted the greatest number of interactions amongst all 171 designed peptides. Hence, it is quite evident that the role of these positions and the amino acids are essential for receptor binding. Our analysis came up with some potential short peptides as shown in Fig. 13. If we consider polar His7 and Glu9, then both polar and non-polar residues have been able to acquire interactions but the presence of aromatic ring and guanidine group at 7th position and carboxylic side chain at 9th position were vital for the interactions. When Phe12 was replaced by arginine the guanidine group again showed interactions with receptor amino acids. Here, Phe12 being a hydrophobic residue was seen to be replaced by hydrophilic residue giving us an interesting quest to the flexibility of this position as far as the nature of the surrounding residues of the receptor. The polar Asp15 being the most consistent interacting residues of the peptide was found to be conserving the interactions when replaced by Trp15 as it interacted with one of the biologically important residue Tyr241.
3.4 Hydration sites role in binding affinity
The WaterMap calculations aided in the discovery of localised hydration hotspots around the binding cavity of the GLP-1 receptor using thermodynamic energetic parameters like enthalpy (ΔH), entropy (− TΔS), and differential binding energy (ΔΔG). The overlapping hydration sites on ligand functional groups could be classified as displaceable (ΔΔG ≫ 0 and ΔH ≫ 0), replaceable (ΔH ≪ 0 and yet ΔΔG ≫ 0 or ≅ 0), and stable (ΔΔG ≪ 0) water molecules from the perspective of drug design based on differential binding energy ΔΔG (Cappel D et al. 2017). In comparison to bulk solvent, the entropy of a hydration site is always unfavourable near the protein (Pearlstein R A et al. 2010). Since the displacement of high-energy hydration sites from the protein binding site is a driving source of binding affinity, a detailed thermodynamic characterization of hydration sites (HSs) is vital for drug design (Beuming T et al.2009). WaterMap calculations were performed using the prepared protein structures 6X18 in this study. Figure illustrates the protein's water map which was produced. For this study, only hydration sites that are within 5 Å of ligand binding sites were analyzed. Hydrophobic areas are recognised as displaceable hydration sites with both ΔΔG and ΔH > > 0 kcal/mol, which can be favourably displaced with sufficient hydrophobic residues of the peptide. It should be emphasised that the displacement of such hydration sites has a significant impact on the peptide's binding. Analysis of the water map of the 6X18 binding site showed ten unstable waters in the vicinity of the peptide, which are displaced by His7, Phe12 and Ser13 side chains. This indicates the role of these residues in increasing the binding affinity of the peptide. The enthalpy (ΔH), entropy (-TΔS) and free energy (ΔΔG) for the ten water molecules, HS36, HS88, HS114, HS119, HS189, HS194, HS224, HS288 and HS315, are given in Table 3.
Table 3
Thermodynamic Properties of Computationally Predicted and Selected Hydration Sites for 6X18
Hydration Site
|
ΔH (kcal/mol)
|
-TΔS (kcal/mol)
|
ΔΔG (kcal/mol)
|
36
|
1.10
|
2.81
|
3.91
|
88
|
1.33
|
1.96
|
3.29
|
114
|
2.82
|
1.52
|
4.34
|
119
|
1.83
|
1.50
|
3.83
|
189
|
1.51
|
1.23
|
2.74
|
194
|
3.34
|
1.17
|
4.51
|
196
|
4.08
|
1.14
|
5.22
|
224
|
3.59
|
0.92
|
4.51
|
288
|
3.46
|
0.88
|
4.34
|
315
|
3.34
|
0.75
|
4.09
|
When the wild type peptide was overlapped on the waterMap hydration sites it was quite evident that His7 displaced three Hydration sites HS36, HS114 and HS119, where as Phe12 displaced HS194 and Ser13 displaced HS315. (See Fig. 14) There were other five hydration sites as well, which where near to other residues of the peptide as shown in Figure. The analysis of the environment of hydration sites showed that they were surrounded by neutral residues like Tyr152, 148, hydrophilic residues like Arg 380, 190, and hydrophobic residues like Trp 306. There are also waters with ΔΔG < < 0, which are difficult to displace or replace but can make water-bridged interactions with proteins. The water molecules like HOH513, HOH518, HOH105, HOH102 were observed to be near unstable hydration sites indicating the importance of these water molecules when mediated interaction with peptides and amino acid of target protein. Also, the most stable hydration sites with ΔG < < 0 (shown as bright green balls in Fig. 14) are a part of the conserved regions of the protein and should be avoided while designing molecules. In 6X18 there is a region in the binding site which can be further explored using the thermodynamic properties of the hydration sites for designing novel peptides. There are a few more hydration sites that can be considered for replacement as per the given thermodynamic parameters. It was observed that many of the seven screened peptides had significant overlap with the reference ligand, and a few of them showed a more significant binding affinity in terms of the calculated parameters.
Therefore, 37 selected peptides were analysed it was observed that 22 peptides have water mediated interactions. Based on vital residue docking interactions, water mediated H-bond interactions and hydration site superimposition we finalised 7 peptides that could be tested experimentally to check their activity (see Fig. 15).
Also, from the seven selected peptides, we designed a peptide with 9 amino acids taking one promising residue from each position (indicated in red Fig. 15). This molecule was docked and superimposed on the waterMap 6X18 to indicate the position hydration sites of the protein in the vicinity of the peptide. (See Fig. 16)