Journal of Biomolecular Research & Therapeutics

Conformationally constrained peptides are very good subjects for investigations, since the provided modifications make the structure more rigid. It helps in studies of active conformations structures in solution. Widely used cis peptide bond constraints include: N-metylated residues [1], double bonds [2], or 1,2,3 triazole [3]. But the most common cis amide bond surrogate is 1,5-disubstituted tetrazole [4,5]. This mimic was successfully introduced to among others bradykinin [6], CCK-B receptor ligands [7], somatostatin [8], enkephalins [9], TRH analogues [10] or scyliorhinin I [11], allowing structural studies of bioactive conformations.

The object of this study Scyliorhinin II (ScyII) was isolated from the dogfish gut in 1986 by Conlon et al. [12]. It is a tachykinin peptide which displays selective agonistic activity towards the NK-3 tachykinin receptor [13]. All tachykinin receptors are of similar sequence and belong to the family of G-protein coupled receptors. Their structure is based on heptahelical structure of rhodopsin [14]. The wide range of physiological activity of tachykinin peptides is caused by their short backbone and linearity [15]. Because of these features, they can easily adopt bioactive conformation in contact with the receptor. Scyliorhinin II is one of the biggest tachykinin peptides. Furthermore, there is a disulfide bridge which is rare structural element among all naturally occuring tachykinins. The amino acid sequence of this peptide is as follows:

Ser¹-Pro-Ser-Asn-Ser-Lys-Cys(&)-Pro-Asp-Gly-Pro-Asp-Cys(&)- Phe-Val-Gly-Leu-Met¹⁸

Literature data [1,16-18] describes selective agonists for NK-3 tachykinin receptor as ones which prefer to adopt α-helical conformation. Our previous studies showed that ScyII does not adopt any particular secondary structure in the solution. Moreover, we detected the existence of cis/trans equilibrium involving three residues of ScyII [19].

Additionally, as reported [1,20,21], Gly16 plays an important role in biological activity and three-dimensional structure of this peptide. Taking above into account, we decided to synthesize two restrained analogues of ScyII. We introduced a tetrazole ring as a surrogate for the cis peptide bond between positions 11 and 12 ([¹¹Ψ¹²(CN₄])]ScyII) and 15 and 16 ([¹⁵Ψ¹⁶(CN₄])]ScyII). In this paper, we describe total conformational analysis of [¹¹Ψ¹²(CN₄)]ScyII and [¹⁵Ψ¹⁶(CN₄)]ScyII molecules in DMSO-d₆ using NMR spectroscopy in conjunction with restrained molecular dynamics calculations. We present our results as a set of low energy conformations and discuss them in terms of structural features in comparison to ScyII and its other analogues.

Peptide synthesis

Both peptides were synthesized according to protocol described previously [11].

NMR experiment

The sample concentrations were approximately 5 mM in DMSO-d₆ for [¹¹Ψ¹²(CN₄)]ScyII and [¹⁵Ψ¹⁶(CN₄)]ScyII. All experiments were carried out on a Varian Unity 500 Plus spectrometer (Varian Instruments USA), operating at 500 MHz resonance frequency at 305 K except for temperature ones, which were measured throughout the temperature range of 295-313 K. The assignment of the proton shifts was made by means of one dimensional proton spectra and two dimensional TOCSY (90 ms) [22], NOESY (400 ms) [23], ROESY (200 ms) [24], and DQFCOSY [25,26]. All NMR data was processed using VNMR 6.1B [27], XEASY 3.1 [28] and CARA 1.2 [29] software.

Vicinal coupling constants

The ³J_NHα coupling constants were extracted from 1D ¹HNMR and 2D DQF-COSY spectra. Due to a great number of overlaping signals in NH region, collecting of ³J_HN-Hα constants was possible in the case of [¹⁵Ψ¹⁶(CN₄)]ScyII only.

NOE effects

All NOE cross-peaks, for peptides studied were picked up in the NOESY spectra. The integration was performed in CARA 1.2.

Conformational calculations

Parameterization of tetrazole groups: Two residues including tetrazole ring were build as Pro[ΨCN₄]Asp and Val[ΨCN₄]Gly. They were modelled using bond lengths, the valence and torsional angles of appropriate residues and compatible molecular segments taken from CSDS database [30]. The partial atomic charges were optimized by fitting the point-charge Coulombic potential to the molecular electrostatic potential calculated using GAMESS program and RHF 6-31 G* wave function [31].

Calculations were performed for two different conformations of every non-standard residue, followed by consecutive averaging the charges over all conformations, as recommended by the RESP protocol [32,33].

Calculations were carried out in CHARMM force field implemented in XPLOR 3.1 package [34]. The starting conformation was set to random. Additionally, NMR-derived constraints for interproton distances, dihedral angles and ω angles of the peptide groups (to keep them in a trans configuration) were added to the target function with force constants: f=50 kcal/mol×Ų, f=50 kcal/mol×rad² and f=500 kcal/ mol×rad², respectively. The chirality of C^α atoms (except for Gly) was fixed to l by imposing a three-fold potential on the improper N-CO-C^α-C^β torsion angles with force constant f=500 kcal/mol×rad².

Assignment of the proton chemical shifts of both peptides was completed using DQF-COSY, TOCSY (Figure 1a and 1b) and NOESY spectra. Spin systems of Val, Leu and Met were identified based on the position of their β, δ and γ protons. Signals of protons of aminoacids joined with a tetrazole group were recognized by the cross-peak between H^α atoms of these residues. Asn 4 protons were possible to identify by means of couplings between H^β and HNδ. All Gly residues were unambiguously identified by their H^α positions. The rest of H^α protons were identified by sequential couplings visible in fingerprint region of NOESY spectra (Figure 2a and 2b). Next using TOCSY spectra, the rest of protons were assigned. Correctness of this assignment was proved by means of DQF-COSY and NOESY. For two residues of [¹⁵Ψ¹⁶(CN₄)]ScyII, we found more than one set of residual proton resonances (Lys6, Val15). It could be connected with either the presence of cis/trans isomerization or flexibility of peptide’s fragments containing these residues. All the chemical shifts are summarized in Table 1a and 1b. In both cases, all peptide bonds were in trans configuration. To obtain interproton distances, 67 and 124 NOE effects were picked for [¹¹Ψ¹²(CN₄)]ScyII and [¹⁵Ψ¹⁶(CN₄)]ScyII, respectively. Obtained NOE pattern for [¹¹Ψ¹²(CN₄)]ScyII and temperature coefficients (Figure 3a) suggested lack of any particular dominant secondary structure element. However d_NN(i,i+2) and d_αN(i,i+2) NOEs of Cys13 and Val15 pointed the existence of two β-turns in regions involving these residues. Moreover, Δδ/ΔT values obtained for these residues indicated involvement of their HN protons in the formation of strong hydrogen bonds. The second peptide [¹⁵Ψ¹⁶(CN₄)]ScyII showed also a rigid structure at the C-terminus. NOE pattern (Figure 3b) suggested existence of two overlapping β-turns in the region Phe14-Met18. One of their determinants was the tetrazole ring between Val15 and Gly16.

Table 1a: The chemical shifts (ppm) of [¹¹Ψ¹²(CN₄)]ScyII in DMSO-d₆ at 305 K.

Table 1b: The chemical shifts (ppm) of [¹⁵Ψ¹⁶(CN₄)]ScyII in DMSO-d₆ at 305 K.

Figure 1: Fingerprint region of TOCSY spectra of Scyliorhinin II analogues in dmso-d6: a) [11Ψ12CN412]ScyII b) [15Ψ16CN4]ScyII.

Figure 2: Fingerprint of NOESY spectra of peptides studied in dmso-d6: a) [11Ψ12CN4]ScyII b) [15Ψ16CN4]ScyII.

Figure 3: Internal intensities of off-diagonal signals in NOESY spectra, the HN temperature coefficients Δ/ΔT (-p.p.b./K) and vicinal coupling constants 3JHN-Hα of ScyII analogues: a) [11Ψ12CN4]ScyII b) [15Ψ16CN4]ScyII.

The vicinal coupling constants indicated extended structure of the peptide’s backbone (most of obtained values are above 8 Hz). Additionally, when comparing values of temperature coefficients, we deducted that the second peptide studied characterized more packed arrangement of the backbone.

Conformational calculations were carried out only for major species because there was too little data to determine minor ones. As a result, we chose ten conformers of the lowest energy from two ensembles of 100 conformations for each of the peptide studied. For obtained structures, we calculated the positions and types of β-turns (Table 2). They pointed the rigid structure of the peptides and were in good agreement with NMR data indicating [¹⁵Ψ¹⁶(CN₄)]ScyII as more rigid and packed than the other peptide.

Table 2: Position and types of β-turns of obtained conformations.

The superposition of all C^α atoms of [¹¹Ψ¹²(CN₄)]ScyII and [¹⁵Ψ¹⁶(CN₄)]ScyII gave RMSDs of 1.778 and 1.869 Å, respectively. In both ensembles of results, we indicated families of conformations with lower RMSD values. They were: the family of 6 conformations with RMSD of 0.878 Å for [¹¹Ψ¹²(CN₄)]ScyII and two families of 4 conformations for [¹⁵Ψ¹⁶(CN₄)]ScyII with RMDSs of 0.597 and 0.555 Å (Figure 4a-4c). Fragments of studied peptides were better defined what was confirmed by the values of corresponding RMSDs. For 10 conformations of [¹¹Ψ¹²(CN₄)]ScyII, superposition of C^α atoms of 7-13 and 12-18 fragments produced RMSDs of 0.792 and 0.546 Å, respectively, whereas for the second peptide, the same fragments gave RMSDs of 1.078 and 1.040 Å. In Figure 5, we showed the comparison of the lowest energy conformations obtained for both ScyII analogues. Analyzing -turns, we could say that IV type β-turn is present in almost all conformations in the regions, which contain tetrazole ring. Positions of other β-turns in each conformational ensemble were similar, but the type.

Figure 4: Superposition of Cα atoms of the obtained conformations: a) family of six conformations of [11Ψ12CN4 12]ScyII, RMDS=0.878 Å, b) and c) families of four conformations of [15Ψ16CN4]ScyII RMSDs equal 0.597 and 0.555 Å respectively. Disulfide bridge is marked yellow and tetrazole red.

Figure 5: Comparison of the lowest energy conformations of [11Ψ12CN4] ScyII (green) and [15Ψ16CN4]ScyII (blue): a) superposition of Cα of the whole backbones, RMSD=4.895 Å; b) superposition of Cα atoms of 3-13 fragments, RMSD=2.934 Å; c) superposition of Cα atoms of 13-18 fragments, RMSD=1.948 Å. Disulfide bridge is marked yellow and tetrazole red.

Studying published data, we have found that 1,5-disubstituted tetrazole is an effective restraint, which allows the peptide to adopt conformation to be recognized by the enzyme [4-6,35-38]. Further, more introduction of this mimic into bradykinin showed that the peptides were able to adopt most conformations of those for native hormone [4-6,35-38].

Not contrary to literature [39], obtained conformations for both ScyII analogues do not adopt any particular secondary structure. Studying the positions of β-turns, we assumed that they were similar to those in [Sar¹⁶]ScyII and [AiB¹⁶]ScyII, but their types were different. Closer analysis of Ramachandran plots obtained for the peptides studied revealed that C-terminus of [¹¹Ψ¹²(CN₄)]ScyII might tend to adopt helical structure, which additionally could be confirmed by d_NN(i,i+2) NOE effect. Such conformation is responsible for biological activity of tachykinin peptides [1], and may be formed in contact with receptor. Introduction of tetrazole between residues 11 and 12 made the C-terminus more rigid and helped expose C-terminal fragment out of the molecule making it more accessible. We met the opposite situation in the case of [¹⁵Ψ¹⁶(CN₄)]ScyII. The IV type β-turn present in the region of tetrazole introduction caused that the Cys13-Met18 fragment resembled the letter U. We assumed that such restriction could disable biological activity of this peptide. Summing up the introduction of tetrazole ring influenced the peptides’ backbones not only locally, but also globally. Furthermore, analyzing the obtained conformations, we could also assume that [¹¹Ψ¹²(CN₄)]ScyII might exhibit biological activity what was connected with its C-terminal fragment structure, which was similar to one obtained by Dike and Cowsik for scyliorhinin II in DPC micelles [40].

This work was supported by a grant from the University of Gdańsk (DS/8290-4-0129-12)