The helical structure of surfactant peptide KL4 when bound to POPC: POPG lipid vesicles

Abstract

KL 4 is a 21-residue peptide employed as a functional mimic of lung surfactant protein B, which successfully lowers surface tension in the alveoli. A mechanistic understanding of how KL 4 affects lipid properties has proven elusive as the secondary structure of KL 4 in lipid preparations has not been determined at high resolution. The sequence of KL 4 is based on the C-terminus of SP-B, a naturally occurring helical protein that binds to lipid interfaces. The spacing of the lysine residues in KL 4 precludes the formation of a canonical amphipathic alpha-helix; qualitative measurements using Raman, CD, and FTIR spectroscopies have given conflicting results as to the secondary structure of the peptide as well as its orientation in the lipid environment. Here, we present a structural model of KL 4 bound to lipid bilayers based on solid state NMR data. Double-quantum correlation experiments employing (13)C-enriched peptides were used to quantitatively determine the backbone torsion angles in KL 4 at several positions. These measurements, coupled with CD experiments, verify the helical nature of KL 4 when bound to lipids, with (phi, psi) angles that differ substantially from common values for alpha-helices of (-60, -45). The average torsion angles found for KL 4 bound to POPC:POPG lipid vesicles are (-105, -30); this deviation from ideal alpha-helical structure allows KL 4 to form an amphipathic helix at the lipid interface.

Figure 1

CD spectra for KL₄ in buffer, in TFE, and when reconstituted with 3:1 POPC: POPG LUVs, at the concentrations indicated.

Figure 2

CPMAS and DQ-filtered spectra for KL₄ ¹³C’-enriched at positions L7 and L8. Left: full ¹³C CPMAS spectra before and after DQ filtering. Right: expansion of ¹³C’ region showing the peptide chemical shifts are constant over the range of conditions used with NMR experiments, peptide: lipid ratios and hydration levels as indicated. Also apparent in the unfiltered spectra are the natural abundance contributions from the peptide, lipids, and rotor materials.

Figure 3

DQ buildup data for KL₄ ¹³C’-enriched at positions as indicated; error bars represent standard deviation for multiple experiments. Also shown are best fit simulations of the data as well as for an ideal α-helix, β-sheet, and mixture of helix and sheet.

Figure 4

Left: 2D-DQ correlation data for KL₄ ¹³C’-enriched at positions L9 and L10 along with a best fit simulation. Closed and open symbols are the real and imaginary data points, respectively; solid and dashed lines correspond to the real and imaginary signal trajectories for the best fit (Φ,Ψ) simulation. Error bars represent standard deviation for multiple experiments. Right: χ² evaluation of simulations with varying Ψ while holding Φ at the value obtained from DQ buildup experiments.

Figure 5

Helical wheel representations of KL₄ assuming a classic α-helix (left) and based on ssNMR measurements (middle). Also shown is a ribbon model of KL₄ based on the structural measurements (right). Green residues are leucines directly characterized by ssNMR measurements; blue and light green residues correspond to lysines and leucines, respectively, for which an average backbone conformation of (-105, -30) is assumed.

Figure 6

Model for the interaction of KL₄ with POPC lipids. Lipid coordinates provided by Scott Fellers, at http://persweb.wabash.edu/facstaff/fellers/ (71).