Effects of guanidinium-phosphate hydrogen bonding on the membrane-bound structure and activity of an arginine-rich membrane peptide from solid-state NMR spectroscopy.

Arginineand lysine-rich cationic peptides and protein domains are found in a wide range of membrane-active proteins such as antimicrobial peptides, cell-penetrating peptides, and voltage-sensing domains of potassium channels. Yet the three-dimensional structures these proteins adopt to enable translocation of the charged residues into the low dielectric milieu of the hydrophobic part of the lipid membrane, despite the free-energy barrier, remain poorly understood. An increasing number of molecular dynamics simulations and experimental studies have suggested the importance of Arg interactions with lipids in membrane protein function. Magic-angle spinning (MAS) solid-state NMR (SSNMR) spectroscopy can provide direct experimental insights into these intriguing questions of energetics and structure. We have recently reported SSNMR distance-constrained guanidinium–phosphate (Gdn–PO4 ) complex formation between the Arg residues of a b-hairpin antimicrobial peptide, PG-1, and the lipid phosphate groups. The existence of these complexes suggests that the Arg residues are neutralized by the phosphate groups to enable transmembrane insertion of the peptide. We hypothesized that the peptide-associated phosphate headgroups transferred to the hydrophobic part of the membrane are responsible for the toroidal pore defects. Such Gdn–PO4 complexes should be stabilized by N H···O=P hydrogen bonds and electrostatic attractions. Herein we test the importance of Gdn–PO4 hydrogen bonding to the structure and activity of PG-1 by dimethylating each guanidinium group, thus reducing the number of N H hydrogen-bond donors (Figure 1). We show that this dimethylation of the Arg groups significantly alters the membrane insertion and activity of PG-1. Figure 2 shows oriented P NMR spectra of palmitoyloleoylphosphatidylcholine/palmitoyloleoylphosphatidylglycerol (POPC/POPG) membranes containing 0–4% of the mutant (Arg-PG-1). Without the peptide, the membranes uniaxially aligned on glass plates and exhibited the expected single peak at approximately 30 ppm without other intensities in the anisotropic chemical shift range. Residual powder intensities indicative of misalignment and a small 0 ppm isotropic peak indicative of toroidal pores are observed with increasing concentrations of Arg-PG-1. Line-shape simulations indicate that the isotropic component in the 4% Arg-PG-1 sample is 20% of the total intensity, much less than the 39% caused by PG-1; thus, dimethylation of the Arg residues reduces membrane disruption. Corroborating the P NMR data are the minimal effective concentrations (MECs) of Arg-PG-1 and PG-1 against Figure 1. a) Structure of dimethylated Arg residue and its bidentate complex with phosphate ions. b) Arg-PG-1 amino acid sequence.