Controlling and fine tuning the physical properties of two identical metal coordination sites in de novo designed three stranded coiled coil peptides

Abstract

Herein we report how de novo designed peptides can be used to investigate whether the position of a metal site along a linear sequence that folds into a three-stranded α-helical coiled coil defines the physical properties of Cd(II) ions in either CdS(3) or CdS(3)O (O-being an exogenous water molecule) coordination environments. Peptides are presented that bind Cd(II) into two identical coordination sites that are located at different topological positions at the interior of these constructs. The peptide GRANDL16PenL19IL23PenL26I binds two Cd(II) as trigonal planar 3-coordinate CdS(3) structures whereas GRANDL12AL16CL26AL30C sequesters two Cd(II) as pseudotetrahedral 4-coordinate CdS(3)O structures. We demonstrate how for the first peptide, having a more rigid structure, the location of the identical binding sites along the linear sequence does not affect the physical properties of the two bound Cd(II). However, the sites are not completely independent as Cd(II) bound to one of the sites ((113)Cd NMR chemical shift of 681 ppm) is perturbed by the metalation state (apo or [Cd(pep)(Hpep)(2)](+) or [Cd(pep)(3)](-)) of the second center ((113)Cd NMR chemical shift of 686 ppm). GRANDL12AL16CL26AL30C shows a completely different behavior. The physical properties of the two bound Cd(II) ions indeed depend on the position of the metal center, having pK(a2) values for the equilibrium [Cd(pep)(Hpep)(2)](+) → [Cd(pep)(3)](-) + 2H(+) (corresponding to deprotonation and coordination of cysteine thiols) that range from 9.9 to 13.9. In addition, the L26AL30C site shows dynamic behavior, which is not observed for the L12AL16C site. These results indicate that for these systems one cannot simply assign a "4-coordinate structure" and assume certain physical properties for that site since important factors such as packing of the adjacent Leu, size of the intended cavity (endo vs exo) and location of the metal site play crucial roles in determining the final properties of the bound Cd(II).

Figure 1

Titration curves obtained by plotting the change in the absorbance at 235 nm as a function of the equivalents of CdCl₂ added to solutions containing 20 µM (grandL16PenL26AL30C)₃ at pH 9.0 (red), 20 µM (grandL12AL16CL26AL30C)₃ at pH 8.6 (blue) and 20 µM (grandL16PenL19IL23PenL26I)₃ at pH 9.5 (green). All the curves plateau at 2 ± 0.14 equivalents of CdCl₂ per equivalent of peptide trimer.

Figure 2

pH dependence of the binding of 2 equivalents of CdCl₂ to 20 µM (grandL16PenL26AL30C)₃ (red), (grandL12AL16CL26AL30C)₃ (blue), (grandL16PenL19IL23PenL26I)₃ (green), (grandL12AL16CL26AL30CL33I)₃ (purple) and of 1 equivalent of Cd(II) to 20 µM (grandL26AL30C)₃ (black), (grandL12AL16C)₃ (orange) and (grandL16Pen)₃ (magenta). UV/Vis absorbance due to LMCT band at 235 nm was monitored during the course of titration and is plotted as normalized absorbance vs. pH.

Figure 3

¹¹³Cd NMR spectra of 3.0 mM [Cd(II)]¹⁶[Cd(II)(H₂O)]³⁰(grandL16PenL26AL30C)₃²⁻ at pH 9.6 (red), 3.4 mM [Cd(II)]₂(grandL16PenL19IL23PenL26I)₃²⁻ at pH 9.5 (green) and 3.3 mM [Cd(II)(H₂O)]₂(grandL12AL16CL26AL30C)₃²⁻ at pH 8.5 (blue). Ball and stick models show the coordination environments of 3- and 4-coordinate Cd(II) complexes. Yellow spheres represent S atoms of Cys/Pen with the green and red spheres representing Cd(II) ion and water molecule, respectively.

Figure 4

¹¹³Cd NMR spectra of solutions containing (A) 3.0 mM (grandL16PenL19IL23PenL26I)₃ and 2 equivalents of ¹¹³Cd(NO₃)₂ at different pH values, and (B) 3.3 mM (grandL16PenL19IL23PenL26I)₃ loaded with 1 and 2 equivalents of ¹¹³Cd(NO₃)₂ at pH 9.5.

Figure 5

¹¹³Cd NMR spectra of solutions containing 3.3 mM (grandL12AL16CL26AL30C)₃ and 2 equivalents of ¹¹³Cd(NO₃)₂ at different pH values. (The most downfield peak marked by star is an impurity).

Figure 6

^111mCd PAC spectra of the different GRAND peptides [Fourier transform: experimental data (thin line) and fits (bold faced line) are shown overlaid]. All the samples contained 20 mM appropriate buffer and 250 – 300 µM peptide. A) grandL26AL30C, 1/12 eq Cd(II), pH 7.0; B) grandL26AL30C, 1/12 eq Cd(II), pH 9.1; C) grandL16PenL26AL30C, 1/12 eq Cd(II), pH 6.5; D) grandL16PenL26AL30C, 1/12 eq Cd(II), pH 9.3; E) grandL16PenL26AL30C, 1.85/3 eq Cd(II), pH 9.3; F) grandL16PenL19IL23PenL26I, 1.85/3 eq Cd(II), pH 9.2; G) grandL12AL16CL26AL30C, 1.85/3 eq Cd(II), pH 8.7.

Figure 7

Sections of ¹H-¹H NOESY spectra of 3.25 mM (grandL12AL16CL26AL30C)₃ as a function of added equivalents of Cd(II) at pH 6.0. Peak at 7.91 ppm corresponds to H^N of E21 and other peaks displayed in gray correspond to interesidue NOEs (H^N_i-H^β_i+1) (see text for full description). Colored peaks correspond to intraresidue NOEs of the amide protons (H^N, vertical axis) and the β methylene protons (H^β, horizontal axis) of Cys: a) H^N₁₆-H^β₁₆ for grandL12AL16CL26AL30C; b) H^N₁₆-H^β₁₆ for [Cd(II)(H₂O)]¹⁶[apo]³⁰(grandL12A16L16CL26AL30C)⁻; and c) H^N₃₀-H^β₃₀ for grandL12AL16CL26AL30C.