Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 290 (29), 18156-72

Hydroxyproline-induced Helical Disruption in Conantokin Rl-B Affects Subunit-selective Antagonistic Activities Toward Ion Channels of N-Methyl-d-aspartate Receptors

Affiliations

Hydroxyproline-induced Helical Disruption in Conantokin Rl-B Affects Subunit-selective Antagonistic Activities Toward Ion Channels of N-Methyl-d-aspartate Receptors

Shailaja Kunda et al. J Biol Chem.

Abstract

Conantokins are ~20-amino acid peptides present in predatory marine snail venoms that function as allosteric antagonists of ion channels of the N-methyl-d-aspartate receptor (NMDAR). These peptides possess a high percentage of post-/co-translationally modified amino acids, particularly γ-carboxyglutamate (Gla). Appropriately spaced Gla residues allow binding of functional divalent cations, which induces end-to-end α-helices in many conantokins. A smaller number of these peptides additionally contain 4-hydroxyproline (Hyp). Hyp should prevent adoption of the metal ion-induced full α-helix, with unknown functional consequences. To address this disparity, as well as the role of Hyp in conantokins, we have solved the high resolution three-dimensional solution structure of a Gla/Hyp-containing 18-residue conantokin, conRl-B, by high field NMR spectroscopy. We show that Hyp(10) disrupts only a small region of the α-helix of the Mn(2+)·peptide complex, which displays cation-induced α-helices on each terminus of the peptide. The function of conRl-B was examined by measuring its inhibition of NMDA/Gly-mediated current through NMDAR ion channels in mouse cortical neurons. The conRl-B displays high inhibitory selectivity for subclasses of NMDARs that contain the functionally important GluN2B subunit. Replacement of Hyp(10) with N(8)Q results in a Mg(2+)-complexed end-to-end α-helix, accompanied by attenuation of NMDAR inhibitory activity. However, replacement of Hyp(10) with Pro(10) allowed the resulting peptide to retain its inhibitory property but diminished its GluN2B specificity. Thus, these modified amino acids, in specific peptide backbones, play critical roles in their subunit-selective inhibition of NMDAR ion channels, a finding that can be employed to design NMDAR antagonists that function at ion channels of distinct NMDAR subclasses.

Keywords: calcium channel; electrophysiology; neurobiology; peptide chemical synthesis; peptide conformation.

Figures

FIGURE 1.
FIGURE 1.
ConRl-B and derived mutants display Mg2+-dependent α-helicity. Scans were from 200 to 250 nm of Mg2+/conG or conRl-B ± MgCl2 (A), apo-conG or conG[▾O10] ± MgCl2 (B), and Mg2+/conG or conRl-B[ΔK8AO▾N8Q] ± MgCl2 (C). The peptide concentrations were 100 μm, and the MgCl2 concentrations were 2 mm, when present. The buffer was 10 mm HEPES, pH 7, at 25 °C. Each curve represents an average of three scans, and the percent helicity is determined as a percent of molar ellipticity of −22,081 degrees cm2 dmol−1, which is observed for conG + 2 mm MgCl2 and represents 100% α-helicity of these classes of peptides.
FIGURE 2.
FIGURE 2.
Fingerprint region of the two-dimensional NOESY spectra contributes toward solving the NMR-derived structure of conantokins saturated with Mg2+. αHN-αCH fingerprint region of the two-dimensional NOESY spectra (mixing time 200 ms) for conRl-B (A), conG[▾O10] (B), and conRl-B[ΔK8AO▾N8Q] (C) is shown. Connectivities are only shown for αN (i,i +1).
FIGURE 3.
FIGURE 3.
Schematic summary of coupling constants, sequential and medium range NOE connectivities, and amide exchange. The peptides represented are conRl-B (A), conG [▾O10] (B), and conRl-B[ΔK8AO▾N8Q] (C). The thickness of the squares correlates with weak, medium, and strong intensity NOE cross-peaks.
FIGURE 4.
FIGURE 4.
TALOS+ graphics of secondary structural predictions from NMR peptide backbone chemical shifts. The random coil interface order parameter (RCI-S2) (A–D, top row) and the ANN-based secondary structure (A–D, bottom row) for each Mg2+·peptide complex, as predicted from input of 13C, 1H, and 15N backbone chemical shifts into TALOS+. Green (top row) indicates chain flexibility and red bars (bottom row) represents a prediction of α-helix for each residue. The height of the bars indicates the probability of the residue being part of a helix.
FIGURE 5.
FIGURE 5.
Energy-minimized structures display differences in the α-helix backbone due to the insertion of Hyp10. Ensembles of 20 energy-minimized structures of conRl-B (A), conG[▾O10] (B), conRl-B [ΔK8AO▾N8Q] (C), and conG (D) are shown as all-atom stick models of backbone and side chains (top), and with backbone only (bottom). The individual conformations were superimposed on backbone Cα atoms. The Gla residues are colored in red, Hyp residues in green, Mg2+ in magenta, and other residues in black.
FIGURE 6.
FIGURE 6.
Final three-dimensional solution structures for conRl-B. Structures were generated for conRl-B and con Rl-B-derived mutants and displayed the kink caused by Hyp10. Average NMR solution structures are shown of Mg2+·conRl-B (A), Mg2+/conG[▾O10] (B), Mg2+·conRl-B [ΔK8AO▾N8Q] (C), and Mg2+/conG (D), respectively. The Gla residues are labeled and shown in stick models.
FIGURE 7.
FIGURE 7.
Subtle difference in the structure for conRl-B at a lower temperature suggests a relatively more stable C-terminal α-helix. The solution structures of conRl-B (A) and conG[▾O10] (B) at 25 °C (left) are compared with the structures at 5 °C (right) in the Mg2+-bound complexes. The backbones of the Gla residues are labeled and shown as sticks.
FIGURE 8.
FIGURE 8.
Charged surfaces of Mg2+/conantokins. Electrostatic grid maps of lowest energy structures of Mg2+·conRl-B (A), Mg2+/conG[▾O10] (B), Mg2+·conRl-B[ΔK8AO▾N8Q] (C), and Mg2+/conG (D) are modeled by AutoDock. The most negatively charged areas are colored red, and the most positively charged areas are colored blue.
FIGURE 9.
FIGURE 9.
Binding of Mg2+ contributes to the heat changes of conRl-B derived mutants measured by isothermal calorimetry. A and B, top, incremental heat changes accompanying titrations with Mg2+ at 25 °C. A and B, bottom, heat/mol after Mg2+ addition are plotted against the molar ratio of Mg2+/peptide. The lines are best fit by a sequential binding model. In each case, a series of 0.5-μl injections of Mg2+ was added to the cell containing 0.3 mm conG[▾O10] (A) or conRl-B[ΔK8AO▾N8Q] (B) in 10 mm HEPES, pH 7. The stock concentration of Mg2+ is 9 mm (A) and 6 mm (B). Each reading was taken over a 60-ms interval and was allowed to equilibrate for 120 ms.
FIGURE 10.
FIGURE 10.
Backbone amide protons of Mg2+·conRl-B undergoes fast solvent exchange. 1H amide proton exchange rates of Mg2+·conRl-B (A), Mg2+/conG (B), Mg2+/conG[▾O10] (C), and Mg2+·conRl-B [ΔK8AO▾N8Q] (D) at 5 °C are shown in the presence of 40mm Mg2+. The amide regions of the 1H NMR spectrum are shown at increasing times after dissolving the peptide in 2H2O. Spectra shown from bottom to top were collected at 4, 24, 44, 64, 84, and 124 min. Assignments of the resonances are indicated above each peak. E, even 20 h after initial transfer of 2H2O, the amide backbone protons in the C-terminal Mg2+/conG and Mg2+·conRl-B [ΔK8AO▾N8Q] remain more stable than the backbone amide protons of the Hyp10-containing peptides, Mg2+·conRl-B and Mg2+/conG[▾O10]. The conantokin concentrations were 2 mm, and the Mg2+ concentrations were 40 mm.
FIGURE 11.
FIGURE 11.
Peptide coordination ligands for Mg2+ based on molecular dynamics simulations. A 200-ns MD snapshot of the Mg2+·conantokin complexes displaying the N-terminal (left) and C-terminal (right) conantokin-binding sites for Mg2+, without peptide backbone restraints. Mg2+·conRl-B (A), Mg2+·conRl-B [ΔK8AO▾N8Q] (B), Mg2+/conG (C), Mg2+/conG [▾O10] (D), and MD simulation of Mg2+/conG (E) were done under the conditions of A–D, except for a 1 kcal/mol/Å2 restraint on the stability of the α-helical backbone, which was sufficient to preserve the α-helical backbone.
FIGURE 12.
FIGURE 12.
Electrophysiological recordings of current influx into mouse cortical neurons. Representative traces of the effect of conRl-B on NMDA/Gly-dependent inward currents in WT (A), GluN2A−/− (B), and GluN2B−/− (C) adult mouse cortical neurons are shown. The cells were first stimulated for 3 s with 50 μm NMDA, 10 μm Gly, 1 μm TTX, 0.5 μm strychnine to determine the maximum current (black lines) and then preincubated for 1 min with 10 μm conRl-B, followed by restimulation with a solution of 10 μm conRl-B, 50 μm NMDA, 10 μm Gly for 3 s (red lines). The ratio of the maximum currents ± conRl-B obtained was used to determine the % inhibition for all conantokin specified on the graph with WT (D), GluN2A−/− (E), and GluN2B−/− (F) mouse cortical neurons.

Similar articles

See all similar articles

Cited by 1 article

Publication types

MeSH terms

Associated data

LinkOut - more resources

Feedback